16nm FinFET | KitGuru https://www.kitguru.net KitGuru.net - Tech News | Hardware News | Hardware Reviews | IOS | Mobile | Gaming | Graphics Cards Mon, 08 Jan 2018 20:02:43 +0000 en-US hourly 1 https://wordpress.org/?v=6.4.3 https://www.kitguru.net/wp-content/uploads/2021/06/cropped-KITGURU-Light-Background-SQUARE2-32x32.png 16nm FinFET | KitGuru https://www.kitguru.net 32 32 PNY GTX 1080 Ti XLR8 OC Gaming 11GB Review https://www.kitguru.net/components/graphic-cards/ryan-martin/pny-gtx-1080-ti-xlr8-oc-gaming-11gb-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/pny-gtx-1080-ti-xlr8-oc-gaming-11gb-review/#comments Mon, 07 Aug 2017 07:59:56 +0000 https://www.kitguru.net/?p=342779 PNY goes "back to basics" with the XLR8 OC Gaming GTX 1080 Ti.

The post PNY GTX 1080 Ti XLR8 OC Gaming 11GB Review first appeared on KitGuru.]]>
The latest rumours surrounding AMD's upcoming Vega GPUs point towards GTX 1070- or 1080-class performance. This means the GTX 1080 Ti will likely remain the best choice in the upper echelons of the consumer graphics card market. KitGuru has already assessed a wide range of GTX 1080 Tis and the PNY XLR8 OC Gaming is another overclocked and custom-cooled solution for the serious gamer to ponder.

PNY's interpretation of Nvidia's GTX 1080 Ti GPU is pure and simple, casting aside many of the advanced or unnecessary features, depending on your perspective, that have become common place on top-tier GPUs. The XLR8 OC Gaming's strategy is a dual slot and triple-fan cooling solution, a full-cover aluminium backplate and factory overclock.

That means, to clarify, there's no RGB lighting, no passive fan mode, no onboard fan headers, no BIOS switches, no customisable decals…you get the idea. This simplistic approach may find appeal among certain audiences, particularly those unimpressed by the RGB lighting “fad” or those not wanting to overload their motherboard with a monstrous triple-slot and air-cooled graphics card.

PNY's factory overclocks aren't particularly adventurous either; a boost clock of 1645 MHz is meagre for an overclocked SKU and the stock memory reflects PNY's conservative strategy with this product.

GPU  Nvidia GTX 1080 Ti Founders Edition  PNY GTX 1080 Ti Blower Edition  PNY GTX 1080 Ti XLR8 OC Gaming
Base Clock
 1480 MHz  1480 MHz  1531 MHz
GPU Boost Clock  1582 MHz  1582 MHz 1645 MHz
Memory Clock Effective
 11008MHz  11008MHz   11008MHz
Memory Bandwidth  484 GB/s  484 GB/s   484 GB/s
Price (£)  £689 / $699  £TBC / $740 £TBC / $735

It's certainly not a graphics card to excite the senses or push the boundaries so PNY has to impress with cooling, noise and, most importantly, pricing.

Be sure to check out our sponsors store EKWB here

The post PNY GTX 1080 Ti XLR8 OC Gaming 11GB Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/pny-gtx-1080-ti-xlr8-oc-gaming-11gb-review/feed/ 4
MSI GTX 1080 Ti Gaming X 11GB Review https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-ti-gaming-x-11gb-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-ti-gaming-x-11gb-review/#comments Wed, 19 Jul 2017 07:21:50 +0000 https://www.kitguru.net/?p=338019 If red and black is still your thing, then MSI's GTX 1080 Ti Gaming X could be for you.

The post MSI GTX 1080 Ti Gaming X 11GB Review first appeared on KitGuru.]]>
The GPU market is in a precarious state due to the impact of cryptocurrency mining on supply and demand. Fortunately for gamers, Nvidia's GTX 1080 Ti seems to have come through unscathed – prices are close to normal and stock is still widely available. KitGuru has already assessed a number of impressive GTX 1080 Ti offerings from the likes of ASUS, Gigabyte, MSI and Palit and today we are analysing the MSI Gaming X variant.

The Gaming X series from MSI is frequently a hot seller on etailer websites and a highly-regarded graphics cards brand across the PC industry, after testing the MSI GTX 1080 Gaming X Plus it is easy to see why. Nvidia's GTX 1080 Ti remains the best high-end option for PC Gamers with deep pockets and third party options like this one from MSI offer the best range of cooling, features and noise, or lack of it.

The formula with MSI's GTX 1080 Ti Gaming X isn't anything unusual. MSI has taken Nvidia's GTX 1080 Ti GPU, overclocked it to 1,658MHz on the core (up from 1,582MHz) and assembled it onto a reworked PCB design. MSI overhauled the cooling with its TwinFrozr cooler which is almost-but-not-quite 2 slots in size. Dimensions have expanded over the MSI GTX 1080 Gaming X in order to account for the increased heat output of the GTX 1080 Ti GPU.

Styling is unquestionably sharp although the red and black colour scheme is beginning to feel a little tiresome. Gamers who haven't settled on a final colour scheme for their build may well end up building their system around a graphics card like this. It's an eye-catching centrepiece for any build although MSI fans looking for something a little more exclusive and flexible may do well to consider the Lightning Z instead.

Clock speeds are undoubtedly conservative in the wider market place but MSI has been cautious in order to preserve the superiority of certain products in its own GTX 1080 Ti portfolio. The Gaming X sits below the Lightning range so the clock speeds of 1,544MHz core and 1,657MHz boost make sense.

GPU  MSI GTX 1080 Ti Lighting Z MSI GTX 1080 Ti Sea Hawk X MSI GTX 1080 Ti Gaming X MSI GTX 1080 Ti Duke OC
Base Clock
 1582 (1607) MHz  1544 (1569) MHz  1544 (1569) MHz 1531 MHz
GPU Boost Clock  1695 (1721) MHz  1657 (1683) MHz  1657 (1683) MHz  1645 MHz
Memory Clock Effective
11124 MHz  11016 (11124) MHz  11016 (11124) MHz  11016 MHz
Memory Bandwidth  489 GB/s  484 (489) GB/s  484 (489) GB/s 484 GB/s
Price (£)  £TBC $TBC  £900 ($800)  £775  ($820) £TBC ($750)

Gamers looking for more speed in a simple click can achieve this with the MSI Gaming App. An OC mode can be enabled which boosts frequencies to 1,569 (1683) MHz on the core and 11,124 MHz on the memory. The price to pay for MSI's Gaming X solution, approximately £775, is inevitably a decent chunk above bargain-basement offerings that start from £675.

Let's see how MSI's GTX 1080 Ti Gaming X shapes up through our suite of tests and whether it justifies that extra cost.

Be sure to check out our sponsors store EKWB here

The post MSI GTX 1080 Ti Gaming X 11GB Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-ti-gaming-x-11gb-review/feed/ 12
MSI GTX 1080 Ti Lightning Z 11GB Review https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-ti-lightning-z-11gb-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-ti-lightning-z-11gb-review/#comments Sat, 08 Jul 2017 18:06:41 +0000 https://www.kitguru.net/?p=339872 Lightning strikes again, we test MSI's latest flagship graphics card.

The post MSI GTX 1080 Ti Lightning Z 11GB Review first appeared on KitGuru.]]>
Water cooling is more popular than ever on high-performance graphics cards like the GTX 1080 Ti, solutions include MSI's Sea Hawk, ASUS Poseidon and Gigabyte Aorus Waterforce. Yet in this high performance market is there still room for the bulky and over-engineered air-cooled option? MSI certainly thinks so with its latest flagship the GTX 1080 Ti Lightning Z.

MSI's Lightning brand is well renowned among the enthusiast community and Lightning products of the past always stood out with a striking yellow colour scheme. Market trends have since shifted and now MSI's Lightning brand is more about features than a particular colour scheme.

Enter the MSI GTX 1080 Ti Lightning Z, a rather sophisticated and strikingly well-equipped graphics card for the high-end market. MSI has pushed the aesthetic focus towards neutrality, enhancing things with a fully fledged RGB lighting system that goes well beyond a mere RGB logo like most graphics card vendors deliver.

There are changes in the cooling department – not least the surprise that MSI is ditching its trusty dual-fan Twin Frozr solution of the past for a triple-fan implementation. MSI, like Gigabyte with its Aorus GTX 1080 Ti, has opted for the 2.5 slot design deciding that outright cooling capability should not be hampered by the need to keep to a seemingly arbitrary size constraint.

MSI has delivered a quirky innovation by having a fully fledged copper heatpipe built into the backplate to provide further cooling potential. There are also a number of heat pipes in close proximity to PCB components which MSI is calling “Close Quarters” heat pipe cooling.

The real impressive characteristic of the Lightning Z is not necessarily the cooling solution but what's going on beneath that. The PCB is engineered for extreme overclocking scenarios with a 14 phase power solution capable of delivering abundant power even when under LN2 cooling and extreme voltages.

An LN2 switch, voltage check points and a number of temperature sensors for the GPU core, memory and VRMs make the Lightning Z an overclocker's dream. MSI is keeping the production run exclusive as only 3000 units will be made, though the vast majority of those will make their way into the hands of consumers – rather than extreme overclockers.

Factory clock speeds are aggressive but not market leading, 1582MHz core boosting up to 1695MHz with a slight overclock to the memory running at 11124MHz.

GPU  MSI GTX 1080 Ti Lighting Z MSI GTX 1080 Ti Sea Hawk X MSI GTX 1080 Ti Gaming X MSI GTX 1080 Ti Duke OC
Base Clock
 1582 (1607) MHz  1544 (1569) MHz  1544 (1569) MHz 1531 MHz
GPU Boost Clock  1695 (1721) MHz  1657 (1683) MHz  1657 (1683) MHz  1645 MHz
Memory Clock Effective
11124 MHz  11016 (11124) MHz  11016 (11124) MHz  11016 MHz
Memory Bandwidth  489 GB/s  484 (489) GB/s  484 (489) GB/s 484 GB/s
Price (£)  £TBC $TBC
 £900 ($800)  £775  ($820) £TBC ($750)

An optional “Lightning” mode can be toggled from within MSI's software that boosts the GPU clock to 1721MHz but there's more frequency than that to be found for the adventurous overclocker. On MSI's other graphics cards that software-triggered faster operational mode is usually called “OC” mode, the OC mode speeds of other MSI graphics cards are tabulated above.

Be sure to check out our sponsors store EKWB here

The post MSI GTX 1080 Ti Lightning Z 11GB Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-ti-lightning-z-11gb-review/feed/ 12
Palit GTX 1080 Ti Super JetStream 11GB Review https://www.kitguru.net/components/graphic-cards/ryan-martin/palit-gtx-1080-ti-super-jetstream-11gb-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/palit-gtx-1080-ti-super-jetstream-11gb-review/#comments Mon, 12 Jun 2017 08:24:57 +0000 https://www.kitguru.net/?p=337615 Palit delivers an affordable and high-end GTX 1080 Ti - consumers should rethink brand loyalties.

The post Palit GTX 1080 Ti Super JetStream 11GB Review first appeared on KitGuru.]]>
Any GTX 1080 Ti will be an expensive purchase but there are graphics card vendors meeting more competitive price points. Palit has always done well, particularly in the UK market, at delivering affordable graphics cards at similar levels of quality and performance to rivals like ASUS or MSI, who usually charge a considerable premium.

Palit's range of JetStream graphics cards have consistently proven themselves worthy at KitGuru. In recent times the Palit GTX 1060, GTX 980 and GTX 970 have all performed strongly with scores over 8 out of 10. Today's test focus, the Palit GTX 1080 Ti Super JetStream, is another iteration of Palit's JetStream range.

Palit offers five variations on the GTX 1080 Ti graphics cards for prospective buyers – Founder's Edition, GameRock, GameRock Premium, JetStream and Super JetStream. This Super JetStream model is not Palit's highest clocked GTX 1080 Ti, that accolade goes to the GameRock Premium, but the Super JetStream is not far off with an out of the box boost clock of 1671MHz, up from the stock Nvidia boost frequency of 1582MHz.

The JetStream cooling solution is a dual-fan implementation, like the MSI GTX 1080 Ti Gaming X, and Palit doesn't skimp on quality or features either. There's still RGB lighting, a backplate, dual BIOS with switches and a semi-passive fan mode. A GTX 1080 Ti that ticks all the boxes and that is competitive in price, what more could a PC Gamer ask for?

GPU  Nvidia GTX 1080 Ti Palit GTX 1080 Ti JetStream Palit GTX 1080 Ti Super JetStream Palit GTX 1080 Ti GameRock Premium
Base Clock
 1480 MHz 1480 MHz  1557 MHz 1594 MHz
GPU Boost Clock  1582 MHz 1582 MHz  1671 MHz  1708 MHz
Memory Clock Effective
 11008 MHz  11008 MHz 11008 MHz 11008 MHz
Memory Bandwidth  484 GB/s  484 GB/s  484 GB/s  484 GB/s
Price (£)  from £650 £660 £680 £670

Be sure to check out our sponsors store EKWB here

The post Palit GTX 1080 Ti Super JetStream 11GB Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/palit-gtx-1080-ti-super-jetstream-11gb-review/feed/ 8
ASUS GTX 1080 ROG Strix OC 11Gbps Review https://www.kitguru.net/components/graphic-cards/ryan-martin/asus-gtx-1080-rog-strix-oc-11gbps-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/asus-gtx-1080-rog-strix-oc-11gbps-review/#comments Thu, 01 Jun 2017 08:29:51 +0000 https://www.kitguru.net/?p=335758 An excellent GTX 1080 in all regards, but encroaching on GTX 1080 Ti prices.

The post ASUS GTX 1080 ROG Strix OC 11Gbps Review first appeared on KitGuru.]]>
Nvidia graphics card partners have refreshed a number of GTX 1080 models to take advantage of faster 11Gbps Micron memory. ASUS were one of the first companies to participate in the new scheme from Nvidia that was announced at its 2017 GDC conference.

ASUS joins MSI in creating a new model specifically for 11Gbps memory modules. Nvidia partners are only allowed to include these uprated memory modules in overclocked models and so ASUS has rolled out the new memory configuration with its Republic of Gamers Strix OC graphics card.

In particular the model on test today is the ROG-STRIX-GTX1080-O8G-11GBPS, which causes the ROG-STRIX-GTX1080-O8G-GAMING to become End of Life (EOL) even if the newer ASUS 11Gbps model has slightly lower out-of-the-box frequencies. The slightly slower-clocked GTX 1080 Advanced Gaming, ROG-STRIX-GTX1080-A8G-GAMING, continues to remain on sale albeit at a more competitive price point and retaining the standard 10Gbps memory configuration.

There are a number of other ASUS GTX 1080 models in the picture, which can get a little confusing – but for all intents and purposes this refreshed model is the best ASUS GTX 1080 currently on sale. Frequencies aside the 11Gbps ASUS card is highly similar to the ASUS GTX 1080 ROG Strix A8G KitGuru tested last year.

This means it has a high performance cooling solution, backplate, full RGB functionality, GPU-controlled 4-pin fan headers and a PCB assembled using the ASUS Auto-Extreme production process.

Out of the box the ASUS graphics card will run in Gaming mode, the default mode, which gives a 1695MHz base and 1835MHz boost, with 11010MHz memory. Installing the ASUS GPU Tweak II software will enable the end user to toggle an OC mode that increases frequencies to 1721MHz base, 1860Mhz boost and 11100MHz memory.

Pricing is higher than most of its rivals with ASUS selling its new 11Gbps GTX 1080 close to the £600 level in the United Kingdom – somewhat exuberant pricing given that GTX 1080 Ti graphics cards currently start from around £650.

GPU  ASUS ROG Strix GeForce GTX 1080 OC 11Gbps  ASUS ROG Strix GeForce GTX 1080 OC (08G) ASUS ROG Strix GeForce GTX 1080 OC (A8G) Nvidia GeForce GTX 1080
Base Clock
 1695 MHz 1759 MHz 1670 MHz  1607 MHz
GPU Boost Clock  1835 MHz 1898 MHz  1809 MHz  1733 MHz
Memory Clock Effective
 11010 MHz 10010 MHz  10010 MHz 10010 MHz
Memory Bandwidth  352.3 GB/s  320.3 GB/s  320.3 GB/s  320.3 GB/s
Price (£)  £600  EOL  £530-580  £480+

Be sure to check out our sponsors store EKWB here

The post ASUS GTX 1080 ROG Strix OC 11Gbps Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/asus-gtx-1080-rog-strix-oc-11gbps-review/feed/ 2
MSI GTX 1080 Gaming X Plus 8GB Review https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-gaming-x-plus-8gb-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-gaming-x-plus-8gb-review/#comments Sat, 27 May 2017 12:56:17 +0000 https://www.kitguru.net/?p=335278 Faster 11Gbps GDDR5X memory for the GTX 1080 is here, we take a look at the MSI Gaming X Plus.

The post MSI GTX 1080 Gaming X Plus 8GB Review first appeared on KitGuru.]]>
The GTX 1080 has been out since mid-2016 and in the wake of a number of high-end releases, including the GTX 1080 Ti  and the Titan Xp, Nvidia has refreshed the GTX 1080 to keep it competitive and in the minds of consumers. MSI's Gaming X Plus is the first of the refreshed GTX 1080s KitGuru has had a chance to assess.

The only noteworthy change between MSI's GTX 1080 Gaming X and the new GTX 1080 Gaming X Plus is a faster memory configuration, made possible by more capable Micron GDDR5X memory chips. The stock memory throughput is now 11 Gbps, instead of 10 Gbps, and that 10 per cent boost to memory bandwidth will give newer GTX 1080s a slight advantage over last year's models as well as making them closer in performance to GTX 1080 Ti graphics cards.

There is little to differentiate these 11 Gbps GTX 1080s from last year's models so expect graphics card vendors to simply re-release their GTX 1080 board designs with faster memory from the factory, slowly phasing down production of units with 10 Gbps memory. Most vendors seem to be choosing to add a new model with some signifier as to the faster memory in the name, MSI has used Plus while ASUS is more to the point including 11 Gbps in the model name.

MSI's GTX 1080 Gaming X Plus is identical to the GTX 1080 Gaming X in terms of its cooling solution, PCB and feature set. That's no bad thing either since our review of that card revealed it to be an excellent graphics card – low noise and temperatures, strong performance and solid build quality.

GPU-Z confirms the 10 per cent boost in memory frequency. MSI has also matched clock speeds between the Gaming X and Gaming X Plus so any performance boost is exclusively from the faster memory, rather than any other factors. A quick scour of UK retailers shows us that pricing is more or less identical between the models, if anything, the older model seems to be more expensive at many retailers which is a peculiar situation.

The overall deal with MSI's new 11 Gbps graphics card is made even sweeter by the fact MSI is running a cashback promotion at the moment, until the end of May, which is giving UK buyers £43 off the price of all of its GTX 1080 graphics cards.

GPU MSI GeForce GTX 1080 Gaming X Plus MSI GeForce GTX 1080 Gaming X Nvidia GeForce GTX 1080
Base Clock
 1683 MHz 1683 MHz 1607 MHz
GPU Boost Clock  1822 MHz  1822 MHz 1733 MHz
Memory Clock Effective
 11010 MHz  10010 MHz 10010 MHz
Memory Bandwidth  352.3 GB/s  320.3 GB/s  320.3 GB/s
Price (£)  £520 £520  £480+

Be sure to check out our sponsors store EKWB here

The post MSI GTX 1080 Gaming X Plus 8GB Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/msi-gtx-1080-gaming-x-plus-8gb-review/feed/ 10
Gigabyte Aorus GTX 1080 Ti Xtreme Edition 11GB Review https://www.kitguru.net/components/graphic-cards/ryan-martin/gigabyte-aorus-gtx-1080-ti-xtreme-edition-11gb-review/ https://www.kitguru.net/components/graphic-cards/ryan-martin/gigabyte-aorus-gtx-1080-ti-xtreme-edition-11gb-review/#comments Fri, 21 Apr 2017 07:10:19 +0000 http://www.kitguru.net/?p=330521 A monstrous graphics card for a monstrous budget. We check out the Aorus GTX 1080 Ti Xtreme Edition.

The post Gigabyte Aorus GTX 1080 Ti Xtreme Edition 11GB Review first appeared on KitGuru.]]>
Nvidia's GTX Titan Xp aside the fastest single GPU graphics card of the moment aimed at gamers and PC enthusiasts is the GTX 1080 Ti. Not content with the default performance Gigabyte, using it's burgeoning Aorus brand, has given the GTX 1080 Ti an “Xtreme Edition” makeover.

The Aorus branding of select Gigabyte graphics cards is a new strategy for 2017 with Gigabyte seeking to develop a prestige brand to better compete with rivals. Presumably, the objective is to cultivate a brand following, loyalty and reputation for excellence similar to the Republic of Gamers (ROG) brand under ASUS.

Irrespective of Gigabyte's strategy and intentions, what Aorus has delivered can only be described as a halo graphics card product. Aorus has opted for one of the most aggressive overclocks on the market for a GTX 1080 Ti at 1607 MHz base clock, up from 1480 MHz on the stock Nvidia model, as well as an overclock to the memory.

Unfortunately for Aorus it has been narrowly beaten to the title of fastest GTX 1080 Ti on sale by Zotac, it's GTX 1080 Ti AMP! Extreme Edition ships at 1645 MHz base clock, 1759 MHz boost clock, from the factory.

Aorus claws some of that frequency back with a special OC mode that can only be enabled by installing the Aorus OC utility. This increases the core frequency to 1632 (1746) MHz and 11448 MHz for the memory. It's not all about clock speeds though as the Aorus GTX 1080 Ti Xtreme Edition has other tricks up its sleeves.

The first of those is bringing back the triple slot and triple fan cooler which is big and brash, but surprisingly effective as we'll see later on. Gigabyte has also embellished the card with a 12 + 2 phase VRM, RGB lighting in multiple locations across the card, a semi-passive fan mode, a special VR HDMI port and Aorus is now using a fully automated PCB assembly process which gives a better overall construction quality. Readers may already be familiar with the ASUS take on this technology which it dubs Auto Extreme.

GPU Nvidia GTX 1080 Ti Gigabyte Aorus GTX 1080 Ti Xtreme Edition  Gigabyte Aorus GTX 1080 Ti Gigabyte GTX 1080 Ti Gaming OC 
Base Clock
 1480 MHz 1607 (1632) MHz  1569 (1594) MHz  1518 (1544) MHz
GPU Boost Clock  1582 MHz  1721 (1746) MHz  1683 (1708) MHz 1632 (1657) MHz
Memory Clock Effective
11010 MHz 11232 (11448) MHz  11010 MHz  11010 MHz
Memory Bandwidth  484 GB/s  494.2 (503.7) GB/s 484 GB/s 484 GB/s
Price (£)  from £700, $700  £810-860, $750  £760-800, $720 £740~760, $700

With a price tag about 10 to 15% higher than an entry level GTX 1080 Ti, the Gigabyte Aorus version locks horns with the ASUS ROG STRIX GTX 1080 Ti that KitGuru recently tested. In terms of price, clock speeds, build quality and overall features the two graphics cards are very similar indeed.

We tested the Aorus GTX 1080 Ti Xtreme Edition in the out-of-the-box gaming mode, not OC mode.

Be sure to check out our sponsors store EKWB here

The post Gigabyte Aorus GTX 1080 Ti Xtreme Edition 11GB Review first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/ryan-martin/gigabyte-aorus-gtx-1080-ti-xtreme-edition-11gb-review/feed/ 21
Xilinx starts to ship 16nm FinFET+ chip ahead of schedule https://www.kitguru.net/components/anton-shilov/xilinx-begins-to-ship-worlds-first-16nm-finfet-chip-ahead-of-schedule/ https://www.kitguru.net/components/anton-shilov/xilinx-begins-to-ship-worlds-first-16nm-finfet-chip-ahead-of-schedule/#respond Sat, 03 Oct 2015 00:48:43 +0000 http://www.kitguru.net/?p=270450 Xilinx, a leading maker of programmable gate arrays as well as special-purpose system-on-chip solutions, this week said that it had made the first customer shipment of its Zynq UltraScale+ processing solution. The chip is made using Taiwan Semiconductor Manufacturing Co.’s second-generation 16nm FinFET process technology (CLN16FF+) with increased performance and lowered power consumption. The Xilinx …

The post Xilinx starts to ship 16nm FinFET+ chip ahead of schedule first appeared on KitGuru.]]>
Xilinx, a leading maker of programmable gate arrays as well as special-purpose system-on-chip solutions, this week said that it had made the first customer shipment of its Zynq UltraScale+ processing solution. The chip is made using Taiwan Semiconductor Manufacturing Co.’s second-generation 16nm FinFET process technology (CLN16FF+) with increased performance and lowered power consumption.

The Xilinx Zynq UltraScale+ all-programmable multi-processor system-on-chip (AP MPSoC) integrates four ARM Cortex-A53 general-purpose cores, two ARM Cortex-R5 real time processing units, and an ARM Mali-400 graphics processor. The MPSoC also includes integrated peripherals, safety and security engines, and advanced power management. The system-on-chip also sports UltraScale programmable logic fabric as well as the new UltraRAM on-chip memory technology. Production of the chip using TSMC’s 16nm FinFET+ fabrication process helped Xilinx to improve its complexity and performance while ensuring low power consumption.

The Zynq UltraScale+ chips were shipped to an undisclosed customer about a quarter ahead of schedule, according to the developer. The company does not reveal details about the volumes, but it looks like Xilinx sold the very first A0 revision chips it got back from the fab to a client after verifying that they are fully-functional and meet performance targets. The ability to sell the A0 revision of a chip indicates perfect design of the Xilinx Zynq UltraScale+ and lack of problems with TSMC’s CLN16FF+.

TSMC’s 16nm FinFET production cycle for a batch of chips is about 90 days. The second batch will be ready around three months after Xilinx places an appropriate order. If Xilinx places an order now, it will get its chips sometimes in Q1 2016.

xilinx_zynq_ultrascale

“TSMC's ongoing collaboration with Xilinx has resulted in the early shipment of a world class 16nm FinFET multiprocessing SoC,” said BJ Woo, vice president of business development at TSMC. “Xilinx and TSMC have clearly demonstrated and delivered industry-leading silicon performance with the lowest power consumption, and highest level of systems integration and intelligence among all programmable logic products to date.”

The Zynq UltraScale+ can be used for a wide variety of applications, including computerized vision systems, such as advanced driver assistance systems (ADAS), surveillance, 5G base stations and other. Thanks to the fact that the MPSoC features general-purpose and highly-parallel graphics processing cores, the Zynq UltraScale+ can process and analyze data as well as make decisions and initiate actuator commands. Xilinx claims that the MPSoC offers a scalable and highly customizable programmable platform that will enable future-proof customer designs.

“The early delivery of our Zynq UltraScale+ MPSoC 16nm device extends our outstanding track record for total execution and absolute quality. We call this a “Three-Peat” – being first to market with leadership products at 28nm, 20nm, and now at 16nm,” said Victor Peng, executive vice president and general manager of the programmable products group at Xilinx.

Discuss on our Facebook page, HERE.

KitGuru Says: This is not the first time when Xilinx begins to ship chips made using TSMC’s leading-edge process technology ahead of all rivals, including Altera. If the company really managed to get a 16nm FinFET design right from the very first revision, this clearly indicates that the company has one of the best engineers in the industry.

The post Xilinx starts to ship 16nm FinFET+ chip ahead of schedule first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/xilinx-begins-to-ship-worlds-first-16nm-finfet-chip-ahead-of-schedule/feed/ 0
TSMC obtains orders from AMD and Nvidia for 16nm FinFET chips https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-obtains-orders-from-amd-and-nvidia-for-16nm-finfet-chips/ https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-obtains-orders-from-amd-and-nvidia-for-16nm-finfet-chips/#comments Tue, 25 Aug 2015 09:24:16 +0000 http://www.kitguru.net/?p=264755 Although Taiwan Semiconductor Manufacturing Co. is behind Samsung Foundry with its first-generation FinFET fabrication technology, which is why it lost orders from Apple and Qualcomm, it will still produce chips using 16nm FinFET manufacturing process for other major customers, including Advanced Micro Devices and Nvidia Corp. Advanced Micro Devices plans to use both 14nm FinFET …

The post TSMC obtains orders from AMD and Nvidia for 16nm FinFET chips first appeared on KitGuru.]]>
Although Taiwan Semiconductor Manufacturing Co. is behind Samsung Foundry with its first-generation FinFET fabrication technology, which is why it lost orders from Apple and Qualcomm, it will still produce chips using 16nm FinFET manufacturing process for other major customers, including Advanced Micro Devices and Nvidia Corp.

Advanced Micro Devices plans to use both 14nm FinFET and 16nm FinFET process technologies for its products due next year. It is believed that the company will order 14nm microprocessors from GlobalFoundries, whereas TSMC will concentrate on manufacturing of next-generation AMD Radeon graphics processing units code-named “Greenland” and other using 16nm FinFET process.

tsmc_wafer_semiconductor_chip_300mm_fab_4

Nvidia Corp. added Samsung Foundry into the list of its manufacturing partners earlier this year, but it will continue to use TSMC’s services for its next-generation GP100 graphics chip based on the “Pascal” architecture.

In addition to AMD and Nvidia, the list of TSMC’s 16nm FinFET clients include Apple, Avago, Broadcom, HiSilicon Technologies, LG Electronics, MediaTek and Xilinx, reports DigiTimes web-site.

Samsung Foundry got orders to produce Apple A9 and Qualcomm Snapdragon 820 system-on-chips using its 14nm fabrication process.

TSMC did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: Despite the fact that TSMC is late to market with its 16nm FinFET manufacturing process, it looks like the company will still get a lot of orders from its customers.

The post TSMC obtains orders from AMD and Nvidia for 16nm FinFET chips first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-obtains-orders-from-amd-and-nvidia-for-16nm-finfet-chips/feed/ 4
Samsung expects graphics cards with 6144-bit bus, 48GB of HBM memory onboard https://www.kitguru.net/components/graphic-cards/anton-shilov/samsung-expects-graphics-cards-with-6144-bit-bus-48gb-of-hbm-memory-onboard/ https://www.kitguru.net/components/graphic-cards/anton-shilov/samsung-expects-graphics-cards-with-6144-bit-bus-48gb-of-hbm-memory-onboard/#comments Fri, 21 Aug 2015 15:56:38 +0000 http://www.kitguru.net/?p=264471 Samsung Electronics indicated that it plans to start volume production of high-bandwidth memory (HBM) next year at the Intel Developer Forum this week. At the trade-show, the company revealed its current vision and expectations concerning HBM. Samsung foresees that eventually high-performance applications (such as GPUs) could feature up to six HBM devices to enable unprecedented …

The post Samsung expects graphics cards with 6144-bit bus, 48GB of HBM memory onboard first appeared on KitGuru.]]>
Samsung Electronics indicated that it plans to start volume production of high-bandwidth memory (HBM) next year at the Intel Developer Forum this week. At the trade-show, the company revealed its current vision and expectations concerning HBM. Samsung foresees that eventually high-performance applications (such as GPUs) could feature up to six HBM devices to enable unprecedented capacities and bandwidth.

As it appears, Samsung plans to skip the first-generation HBM memory and will only manufacture products compliant with the second-generation HBM specification, which offers higher densities and clock-rates. Such approach will let Samsung to address broader market segments with its HBM offerings. Right now HBM can only be used on consumer graphics cards and certain highly-custom products, but HBM2 will enable professional and high-performance computing GPU-based solutions, which require a lot of memory onboard.

Samsung is currently working on multiple HBM packages featuring two (2Hi stack), four (4Hi) and eight (8Hi) 8Gb memory devices on a base logic die with 1024-bit interface, according to a slide that the company demonstrated at the IDF (which was published by ComputerBase.de). Maximum data-rates of Samsung’s HBM products will be 2Gb/s, which will support up to 256GB/s of bandwidth per chip.

samsung_hbm_plans

Samsung believes that HBM memory will enable it to create a variety of chip offerings targeting different market segments. Designers of logic chips (GPUs, APUs, network processors, etc.) will be able to integrate the right amount of HBM controllers into their chips in order to target different applications. At present AMD’s “Fiji” graphics processing unit supports up to four HBM stacks over its 4096-bit interface. Eventually, logic chips could accommodate more HBM controllers and expand interface width to unprecedented 6144-bit, according to Samsung Electronics.

For example, a mainstream graphics card could use just one 2Hi HBM chip to accommodate a 2GB frame-buffer with 256GB/s bandwidth. More advanced graphics adapters for consumers and creative professionals could feature four 2Hi or four 4Hi HBM stacks that provide up to 1TB/s of bandwidth as well as 8GB, 16GB or 32GB of memory. Accelerators for high-performance computing as well as ultra-high-end GPU offerings will support six HBM stacks, thus enabling cards with 12GB, 24GB or 48GB of onboard memory with 1.5TB/s bandwidth.

amd_radeon_fiji_gpu

So far, neither AMD nor Nvidia have demonstrated even hypothetical product implementations (which are used to showcase potential future uses of technologies) of GPU-based solutions featuring six HBM memory chips. Intel Corp.’s Xeon Phi co-processors use HMC [hybrid memory cube] DRAMs as “near memory” and are not expected to support HBM any time soon.

nvidia_pascal_module

Samsung expects HBM memory to be used for consumer graphics cards and high-performance computing accelerators based on AMD “Arctic Islands” and Nvidia “Pascal” graphics processors next year. Sometimes in 2017, network products will also take advantage of the new memory type. Three years from now other applications could employ HBM, according to Samsung.

The world’s largest maker of dynamic random access memory (DRAM) did not unveil any details about its actual HBM chips (capacities, clock-rates, etc.), but expect a family of products with different densities and frequencies.

Samsung did not disclose which process technology it will use to manufacture HBM.

Discuss on our Facebook page, HERE.

KitGuru Says: Samsung is definitely not the first to offer HBM, but it is pretty clear that the company takes the new type of memory very seriously. What is interesting is that the company decided not to give a glimpse into the future of HBM, which indicates that the third-generation HBM is still a work in progress.

The post Samsung expects graphics cards with 6144-bit bus, 48GB of HBM memory onboard first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/samsung-expects-graphics-cards-with-6144-bit-bus-48gb-of-hbm-memory-onboard/feed/ 31
AMD readies three new GPUs: Greenland, Baffin and Ellesmere https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-readies-three-new-gpus-for-2016-greenland-baffin-and-ellesmere/ https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-readies-three-new-gpus-for-2016-greenland-baffin-and-ellesmere/#comments Thu, 20 Aug 2015 21:27:02 +0000 http://www.kitguru.net/?p=264345 In the recent years, Advanced Micro Devices has reduced the amount of new graphics processors it releases per annum, which lead to massive erosion of its market share and revenue. While the company hopes that its latest product lineup will help it to regain some of the lost share and improve earnings, the firm pins …

The post AMD readies three new GPUs: Greenland, Baffin and Ellesmere first appeared on KitGuru.]]>
In the recent years, Advanced Micro Devices has reduced the amount of new graphics processors it releases per annum, which lead to massive erosion of its market share and revenue. While the company hopes that its latest product lineup will help it to regain some of the lost share and improve earnings, the firm pins considerably more hopes on an all-new family of products that will be released in 2016.

AMD’s code-named “Arctic Islands” family of products will include three brand-new chips – “Greenland”, “Baffin” and “Ellesmere” – a source with knowledge of AMD plans said. The “Greenland” will be the new flagship offering for performance enthusiasts, whereas the “Baffin” and “Ellesmere” will target other market segments, such as high-end and mainstream. It is unclear whether the “Arctic Islands” family will take in any existing products, but it is possible that AMD may address certain markets with previous-gen products.

amd_radeon_artwork_angle_new

“Greenland” will be AMD’s first graphics processing unit based on the all-new micro-architecture, which development began a little more than two years ago. While the architecture is currently known as another iteration of GCN, the new ISA [instruction set architecture] will be so considerably different compared to the existing GCN that it has every right to be called “post-GCN”, the source said. It is likely that the “Greenland” will retain layout of the contemporary AMD Radeon graphics processing units, but there will be significant changes in on the deeper level.

The only official thing currently known about the new architecture, which Mark Papermaster, chief technology officer of AMD, calls the next iteration of GCN, is that it is projected to be two times more energy efficient compared to the current GCN. Essentially, this means means major performance enhancements on the ISA level. Thanks to the fact that the “Greenland” graphics processing unit will be made using either 14nm or 16nm FinFET process technology, expect it to feature considerably larger number of stream processors than “Fiji”.

amd_radeon_fiji_gpu

The “Greenland” graphics processor will rely on the second-generation high-bandwidth memory (HBM), so expect ultra-high-end graphics cards and professional solutions with up to 32GB of DRAM onboard with bandwidth of up to 1TB/s. Consumer-class “Greenland”-based products will likely come with 8GB – 16GB of memory. Due to usage of HBM, expect the “Greenland” chip and upcoming graphics cards on its base to resemble the currently available AMD Radeon R9 Fury-series adapters.

The number of transistors inside the “Greenland” as well as its die size are unknown. Since 14nm/16nm FinFET manufacturing technologies have considerably (up to 90 per cent) higher transistor density than contemporary TSMC’s 28nm fabrication process, it is logical to expect that the new flagship product will feature 15 – 18 billion of elements if it retains around 600mm² die size from the “Fiji”.

It is believed that AMD has already taped-out its “Greenland” graphics processing unit and is about to get the first silicon in the coming weeks.

amd_radeon_shop-home-component

Not a lot of information is known about the “Baffin” and the “Ellesmere”. The source stressed that both GPUs are brand-new and will be designed from scratch. Since the “Baffin” and the “Ellesmere” are named after bigger and smaller islands in Canada, it is likely that the former is a mainstream graphics chip with moderate die size, whereas the former is a small entry-level GPU. AMD began to work on “Ellesmere” about a year ago.

Expect AMD to begin talking about its next-generation graphics architecture in the coming months.

AMD did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: As it traditionally happens, everything looks very good on paper. If AMD manages to release three brand new chips within a reasonable amount of time in 2016 and the chips will be competitive against Nvidia’s, then it has all the chances to win back market share. However, one should keep in mind that the “next GCN” or the “post-GCN” will not compete against Nvidia’s “Maxwell”, but will have to compete against Nvidia’s “Pascal”, which promises to be very powerful. As a result, expect 2016 is going to be an interesting year for the GPUs…

The post AMD readies three new GPUs: Greenland, Baffin and Ellesmere first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-readies-three-new-gpus-for-2016-greenland-baffin-and-ellesmere/feed/ 87
TSMC begins shipments of chips made using 16nm FinFET process tech https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-begins-shipments-of-chips-using-16nm-finfet-process-technology/ https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-begins-shipments-of-chips-using-16nm-finfet-process-technology/#comments Fri, 17 Jul 2015 12:26:37 +0000 http://www.kitguru.net/?p=259513 Taiwan Semiconductor Manufacturing Co. announced on Thursday that it had started volume production of chips using its 16nm FinFET manufacturing technology in the second quarter of 2015. The world’s largest contract maker of semiconductors is delivering the first batch of products made at 16nm node to its customers right now, according to chief executive of …

The post TSMC begins shipments of chips made using 16nm FinFET process tech first appeared on KitGuru.]]>
Taiwan Semiconductor Manufacturing Co. announced on Thursday that it had started volume production of chips using its 16nm FinFET manufacturing technology in the second quarter of 2015. The world’s largest contract maker of semiconductors is delivering the first batch of products made at 16nm node to its customers right now, according to chief executive of TSMC. The high-volume ramp of 16nm FinFET tech begins this quarter, on-track with expectations.

“We have begun volume production of 16 FinFET in second quarter,” said C.C. Wei, president and co-CEO of TSMC, during the company’s earnings conference call with investors and financial analysts. “Shipment has started this month. The high volume ramp in third quarter [will be] mostly contributed to revenue in fourth quarter this year.”

Previously TSMC said that high-volume production of chips using its 16nm FinFET (CLN16FF) process technology starts in the third quarter, with meaningful revenue contribution beginning in the Q4 2015. The current announcement confirms that high-volume ramp of TSMC’s first process technology with FinFETs [fin-shaped field-effect transistors] proceeds in accordance with the plans.

tsmc_wafer_semiconductor_chip_300mm_fab

TSMC’s 16nm FinFET (CLN16FF), 16nm FinFET+ (CLN16FF+) and 16nm FinFET compact (CLN16FFC) fabrication processes all rely on metal back end of line (BEOL) interconnect flow of 20nm manufacturing technology. As a result, while the chips produced using TSMC’s 16nm tech cannot boast with leading-edge transistor density and have very high per-transistor costs, TSMC can ensure rather high yields with its new process.

“Since 16nm shares similar metal backend process with 20nm, our 16nm FinFET can benefit a lot from 20nm's learning,” said Mr. Wei. “We have already shipped more than half of a million CLN20SOC wafers by now, so our 16nm’s yield and defect density has been excellent. In fact, our 16nm FinFET has set a new record for progresses made in the defect density reduction.”

tsmc_fab_semiconductor

Because 16nm FinFET fabrication process is more universal than TSMC’s CLN20SOC, which was designed solely for mobile system-on-chips, more customers will use the new manufacturing technology. TSMC is aggressively expanding its 16nm capacities and is also converting some of its 20nm product lines to 16nm. The company expects wafer revenue for 16nm process to account over 20 per cent of its total wafer revenue next year.

“The ramping of our 16nm will be very steep, even steeper than our 20nm,” said Mr. Wei.

apple_iphone_6_white_5

TSMC does not disclose names of its customers, but it is highly likely that the first company to get TSMC’s 16nm chips is Apple. It is possible that the lion’s share of TSMC’s 16nm capacities in Q3 2015 will be used to make Apple’s A9 chips for next-generation iPhone and iPad products. TSMC is expected to deliver the first high-volume batch of 16nm FinFET products in October. Initial next-gen smartphones from Apple will be based on semiconductors produced by Samsung Electronics using its 14nm LPE fabrication process.

Discuss on our Facebook page, HERE.

KitGuru Says: TSMC is over half of a year behind Samsung Foundry with volume production of chips using a FinFET process. Keeping in mind vast manufacturing capacities of TSMC, the company will probably beat Samsung in terms of volumes. However, when it comes to time-to-market, Samsung is clearly ahead of its rival.

The post TSMC begins shipments of chips made using 16nm FinFET process tech first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-begins-shipments-of-chips-using-16nm-finfet-process-technology/feed/ 3
AMD reportedly has priority supply agreement regarding HBM, HBM2 https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-reportedly-has-priority-supply-agreement-regarding-hbm-hbm2/ https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-reportedly-has-priority-supply-agreement-regarding-hbm-hbm2/#comments Tue, 14 Jul 2015 10:55:38 +0000 http://www.kitguru.net/?p=258850 The first-generation high-bandwidth memory provides extreme performance, but is clearly not perfect for high-end graphics cards due to capacity limitations. A good news is that second-gen high-bandwidth memory is on the way and is due in 2016. Advanced Micro Devices may get it faster and in higher volumes than its arch-rival Nvidia Corp. The only …

The post AMD reportedly has priority supply agreement regarding HBM, HBM2 first appeared on KitGuru.]]>
The first-generation high-bandwidth memory provides extreme performance, but is clearly not perfect for high-end graphics cards due to capacity limitations. A good news is that second-gen high-bandwidth memory is on the way and is due in 2016. Advanced Micro Devices may get it faster and in higher volumes than its arch-rival Nvidia Corp. The only question is when exactly HBM2 is set to be available.

HBM2 to solve HBM's problems

One of the key drawbacks of the first-generation high-bandwidth memory (HBM) is its low capacity. At present, it is impossible to install more than 4GB of HBM on a single-GPU graphics card, which was not a good news for AMD’s Radeon R9 Fury-series adapters. SK Hynix, AMD, JEDEC and a number of other industry players are working on the second-gen high-bandwidth memory (HBM 2), which will solve a lot of problems, including capacity-related. Moreover, Advanced Micro Devices may benefit from priority supply agreement it reportedly has with SK Hynix.

sk_hynix_hbm_dram_2

The first-generation HBM (HBM1) stacks four DRAM dies with two 128-bit channels per die on a base logic die, creating a memory device with a 1024-bit interface. Each channel supports 1Gb capacities (2Gb per die), features 8 banks and can operate at 1Gb/s data-rate (1GHz effective DDR clock-rate). As a result, each HBM 4Hi stack (4 high stack) package can provide 1GB capacity and 128GB/s memory bandwidth. The second-generation HBM (HBM2) utilizes 8Gb dies with two 128-bit channels featuring 16 banks and sporting up to 2Gb/s data-rates (2GHz effective DDR frequency). The architecture of the HBM2 will let manufacturers built not only 4Hi stack (4 high stack) packages, but also 2Hi stack and 8Hi stack devices. As a result, memory producers will be able to assemble HBM2 memory chips with up to 8GB capacity (8Hi stack) and up to 256GB/s bandwidth (2Gb/s data rate, 1024-bit bus). Architectural advantages of HBM2 will allow GPU developers to use it not only for ultra-high-end applications with 4096-bit memory bus, but also for adapters that do not require extreme performance.

The industry needs HBM2

Advanced Micro Devices has already announced that it would use HBM memory going forward for a broad range of its ICs [integrated circuits], including graphics processing units and accelerated processing units. Right now, the company is working on its all-new family of graphics processing units based on the next iteration of GCN architecture with increased power efficiency and performance. AMD’s next high-end GPU code-named “Greenland” will be made using 16nm or 14nm FinFET process technology, which will help AMD to considerably increase transistor and stream processor counts compared to the code-named “Fiji” (AMD Radeon R9 Fury-series). As a result, “Greenland” will require more advanced memory and AMD plans to use HBM2 for it.

nvidia_pascal_module

Nvidia will also produce its code-named “Pascal” chips using TSMC's 16nm FinFET process technology and will also dramatically boost transistor count vs. existing graphics processing unit. It is expected that Nvidia's code-named GP100 chip will also need a high-end memory sub-system, presumably based on HBM2. At least, Nvidia officially revealed plans to use “3D memory” with “Pascal”.

Supply agreement

Since AMD has been a key developer of both HBM and HBM2, there is a contract in place that gives AMD priorities in terms of supply, according to WccfTech. While the terms of the agreement are unknown, typically such agreements mean that SK Hynix has to meet AMD’s requirements first and only then ship its chips to other customers. Theoretically, this could give AMD “an edge against its main rival, Nvidia, going into the next generation of GPUs featuring second-generation HBM technology.”

Supply agreements are usually signed in a bid to ensure priority treatment, high volumes and good prices. For example, Apple usually signs supply agreements to ensure that it gets what it needs from its suppliers and at low costs. AMD's share in the GPU market is pretty low these days and it may simply not need a lot of HBM2 chips for its launch next year.

HBM2 is still not quite there yet

Earlier this year SK Hynix demonstrated a 300mm wafer with HBM2 chips on it, which is an indicator that the company is testing such products internally, but this does not give any idea about when the HBM2 is ready for mass production.

At present SK Hynix uses 29nm manufacturing technology to manufacture HBM memory chips. According to roadmap leaked by SemiAccurate some time ago, the company plans to use its 21nm fabrication process to produce HBM2 sometimes in Q1 or Q2 2016. SK Hynix’s transition to 21nm is generally considered to be slow, which means that actual launch timeframe may be slightly delayed.

sk_hynix_tsv_roadmap_hbm

SK Hynix thoroughly lists all of its products – current, future and even some EOLed – in its catalogue that is updated every quarter. It is well known, which of SK Hynix’s chips are available, which will be available in Q1 or Q2 2016 and which will be sampled in Q1 2016. At present, SK Hynix does not list HBM2 anywhere, which may be a matter of trade secrets. However, back in 2014, the company did list first-gen HBM chips as “sampling” in Q3 and as “available” in Q4. At the present time, nothing is known about availability or sampling of HBM2 chips, which may indicate that their availability timeframe is not exactly clear.

AMD and SK Hynix did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: Without any doubts, HBM may have certain teething problems, but nobody is talking about them publicly because of commercial reasons. Before the problems get solved, it is unlikely that SK Hynix or other producers of memory start to manufacture HBM2 in higher volumes.

The post AMD reportedly has priority supply agreement regarding HBM, HBM2 first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-reportedly-has-priority-supply-agreement-regarding-hbm-hbm2/feed/ 13
Xilinx tapes-out world’s first 16nm FinFET+ system-on-chip https://www.kitguru.net/components/anton-shilov/xilinx-tapes-out-worlds-first-16nm-finfet-system-on-chip/ https://www.kitguru.net/components/anton-shilov/xilinx-tapes-out-worlds-first-16nm-finfet-system-on-chip/#respond Wed, 01 Jul 2015 22:14:07 +0000 http://www.kitguru.net/?p=256996 Xilinx, a leading designer of field-programmable gate arrays as well as special-purpose system-on-chip solutions, on Wednesday said that it had taped-out one of the world’s first SoCs to be made using TSMC’s 16nm FinFET+ process technology. The new chip will be used for self-driving vehicles, industrial Internet-of-Things and 5G wireless systems in the coming years. …

The post Xilinx tapes-out world’s first 16nm FinFET+ system-on-chip first appeared on KitGuru.]]>
Xilinx, a leading designer of field-programmable gate arrays as well as special-purpose system-on-chip solutions, on Wednesday said that it had taped-out one of the world’s first SoCs to be made using TSMC’s 16nm FinFET+ process technology. The new chip will be used for self-driving vehicles, industrial Internet-of-Things and 5G wireless systems in the coming years.

The new Xilinx Zynq UltraScale+ all-programmable multi-processor system-on-chip (AP MPSoC) integrates four ARM Cortex-A53 general-purpose cores, two ARM Cortex-R5 real time processing units, and an ARM Mali-400 graphics processing unit. The MPSoC also includes integrated peripherals, safety and security engines, and advanced power management. The system-on-chip also sports UltraScale programmable logic fabric as well as the new UltraRAM on-chip memory technology.

The Zynq UltraScale+ can be used for a wide variety of applications, including computerized vision systems, such as advanced driver assistance systems (ADAS), surveillance and other. Thanks to the fact that the MPSoC features general-purpose and highly-parallel graphics processing cores, the Zynq UltraScale+ can process and analyze data as well as make decisions and initiate actuator commands. Xilinx claims that the MPSoC offers a scalable and highly customizable programmable platform that will ensure future-proof customer designs in the quickly changing ADAS space. Besides, increased compute performance of the Zynq UltraScale+ MPSoC will also support the increased radio and baseband processing requirements of next generation 5G systems.

xilinx_zynq_ultrascale

The tape-out of the chip by Xilinx means that the developer will get the first sample of the Zynq UltraScale+ MPSoC several months from now and will start its mass production of the product using TSMC’s most advanced 16nm FinFET+ process technology sometimes in mid-2016.

“The Zynq UltraScale+ MPSoC provides an ideal mix of software intelligence, hardware optimization, security and safety, and any-to-any connectivity for the next-generation of smart and connected applications,” said Victor Peng, executive vice president and general manager of the programmable products group at Xilinx. “The Zynq UltraScale+ MPSoC has been specifically tailored to meet the unique requirements of next generation embedded vision, including ADAS and the roadmap to autonomous vehicles, industrial-IoT and 5G wireless systems, with applicability to numerous other applications.”

Discuss on our Facebook page, HERE.

KitGuru Says: The tape-out of Xilinx confirms that the first wave of chips to be made using TSMC’s CLN16FF+ manufacturing technology has taped out. It is believed that Nvidia Corp.’s code-named GP100 graphics processing unit – the first GPU to be based on the company’s “Pascal” architecture – is also among the first to be made using TSMC’s 16nm FinFET+ fabrication process.

The post Xilinx tapes-out world’s first 16nm FinFET+ system-on-chip first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/xilinx-tapes-out-worlds-first-16nm-finfet-system-on-chip/feed/ 0
Nvidia: TSMC is our primary partner for 10nm and 16nm chips https://www.kitguru.net/components/anton-shilov/nvidia-tsmc-is-our-primary-partner-for-10nm-and-16nm-chips/ https://www.kitguru.net/components/anton-shilov/nvidia-tsmc-is-our-primary-partner-for-10nm-and-16nm-chips/#comments Fri, 08 May 2015 11:14:25 +0000 http://www.kitguru.net/?p=248571 Being one of the largest fabless designers of semiconductors in the world, Nvidia Corp. is a very special customer of Taiwan Semiconductor Manufacturing Co. For more than a decade, TSMC has been the primary producer of Nvidia graphics processing units and despite of Nvidia’s recent engagement with Samsung Foundry, TSMC will remain the company’s key …

The post Nvidia: TSMC is our primary partner for 10nm and 16nm chips first appeared on KitGuru.]]>
Being one of the largest fabless designers of semiconductors in the world, Nvidia Corp. is a very special customer of Taiwan Semiconductor Manufacturing Co. For more than a decade, TSMC has been the primary producer of Nvidia graphics processing units and despite of Nvidia’s recent engagement with Samsung Foundry, TSMC will remain the company’s key manufacturing partner.

“We are constantly evaluating foundry suppliers,” said Jen-Hsun Huang, chief executive officer of Nvidia, during the company’s quarterly conference call with investors and financial analysts. “We largely purchase from TSMC, the vast majority of our wafers we buy from TSMC. We are in 20nm, we are expecting to ramp 16nm. We are deeply engaged with TSMC for many, many nodes to come, including 10nm.”

semiconductor_wafer_cadence_4645_TSMC-Fab3inr008_78-sized

While TSMC remains the world’s largest contract maker of microelectronics, the company is behind Samsung Electronics with its 16nm manufacturing process that employs fin-shaped field-effect transistors (FinFET). Samsung is already producing chips using its 14nm FinFET (14LPE) fabrication technology in high volume, whereas TSMC plans to start making 16nm FinFET chips in Q3 with meaningful revenue contribution starting in the Q4 2015.

At present it is too late for Nvidia to jump ships from 16nm to 14nm (i.e., from TSMC to Samsung). In a bid to get a 14nm FinFET commercial chip from Samsung Electronics in late Q4 2015, Nvidia would have needed to form a design implementation team of around 100-200 engineers a couple of years ago and start to design its ASIC [application specific integrated circuit] in 2013 at the latest. Nvidia would tape out the chip in late 2014 or early 2015, nine to twelve months before the start of mass production. It is unknown when exactly Nvidia decided to use Samsung as a foundry, hence, the company’s 14nm FinFET roadmap and plans are completely unclear.

Officially, Nvidia claims that it does not necessarily need a bleeding-edge manufacturing process to deliver great products.

“There are just so many ways for us to deliver energy efficiency and performance,” said Mr. Huang. “I would not get too obsessed about the process technology all by itself.”.

tsmc_semiconductor_fab14_production

Nonetheless, process technologies are crucial for Nvidia. If Intel or AMD release their new products two or more quarters ahead of Nvidia, the company’s revenue and market share will decrease. Therefore, it makes a great sense for Nvidia to have two foundry partners. However, since it takes a long time to design a modern chip, decisions regarding manufacturing and process technologies have to be made years before such IC [integrated circuit] hits the market. At present Nvidia seems to be downplaying its relationship with Samsung, but no one knows what may happen in the future.

“But we are always looking at new foundry suppliers, and competition keeps everybody sharp,” said the CEO of Nvidia. “But for all intents and purposes, TSMC is our primary partner.”

Discuss on our Facebook page, HERE.

KitGuru Says: “Looking” at new foundry partners and listing one as a supplier in your filing with the Securities and Exchange Commission are two completely different things. In general, it seems that Nvidia is about to start using Samsung as a foundry partner, but it is completely unclear what exactly Samsung will produce for Nvidia. Perhaps, after learning about TSMC’s 20nm focus in late-2011 – early-2012, the company decided to diversify its wafer suppliers. If this is the case, then Nvidia’s next-gen “Pascal” GPUs as well as Tegra system-on-chips will be made by both Samsung and TSMC. Unfortunately, there are not a lot of facts to prove that theory.

The post Nvidia: TSMC is our primary partner for 10nm and 16nm chips first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/nvidia-tsmc-is-our-primary-partner-for-10nm-and-16nm-chips/feed/ 1
AMD may be developing APUs for TSMC’s 16nm FinFET process technology https://www.kitguru.net/components/anton-shilov/amd-may-be-developing-apus-for-tsmcs-16nm-finfet-process-technology/ https://www.kitguru.net/components/anton-shilov/amd-may-be-developing-apus-for-tsmcs-16nm-finfet-process-technology/#comments Sat, 02 May 2015 06:55:25 +0000 http://www.kitguru.net/?p=247811 Although leaked roadmaps of Advanced Micro Devices indicate that the company’s next year’s accelerated processing units and central processing units will all be made using 14nm FinFET fabrication process at GlobalFoundries, this does not mean that AMD plans to cease using Taiwan Semiconductor Manufacturing Co. as a production partner. For many years AMD has used …

The post AMD may be developing APUs for TSMC’s 16nm FinFET process technology first appeared on KitGuru.]]>
Although leaked roadmaps of Advanced Micro Devices indicate that the company’s next year’s accelerated processing units and central processing units will all be made using 14nm FinFET fabrication process at GlobalFoundries, this does not mean that AMD plans to cease using Taiwan Semiconductor Manufacturing Co. as a production partner.

For many years AMD has used both GlobalFoundries and TSMC as its key manufacturing partners. GlobalFoundries focused on higher-end APUs and CPUs, whereas TSMC produced GPUs, inexpensive APUs as well as semi-custom APUs. Last year AMD expanded its orders to GlobalFoundries with select GPUs and semi-custom system-on-chips for game consoles. Next year all key APUs and CPUs from AMD, including “Summit Ridge”, “Bristol Ridge”, “Basilisk” and “Styx” will be made using 14nm FinFET process technology at GlobalFoundries.

Taiwan Semiconductor Manufacturing Co. will likely continue to produce at least some graphics processors for AMD next year. Besides, a quick look at activities of AMD engineers at LinkedIn reveals that many people are (or, perhaps, were) working on accelerated processing units to be made using 16nm FinFET process technology at TSMC. The engineers do not reveal code-names of projects they are (or were) working on.

amd_apu_beema_mullins_puma_jaguar_x86_fusion

Many fabless chip designers are not confident of manufacturing capacities and yields of products made using various FinFET manufacturing processes. To mitigate risks associated with yields and production capacities, numerous developers create similar chips for different contract makers of semiconductors so to ensure stable supply.

AMD could take the approach and use TSMC as its second source if it needs. However, given that AMD is cash strapped and does not require extremely high volumes of APUs and CPUs, it is unlikely that the company is re-designing any of its PC chips for TSMC’s process technology. What is possible is that some of the company’s engineers are working on certain semi-custom APUs to be made using 16nm FinFET process technology. Perhaps, Microsoft and Sony would like to get more energy-efficient system-on-chips for Xbox One and PlayStation 4 next year and are willing to pay for appropriate redesigns.

AMD did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: While AMD will continue to use TSMC as a manufacturing partner, it is obvious that GlobalFoundries is set to become a key producer for the chip developer. AMD wants to re-use maximum amount of chip building blocks it develops across its product line, which is possible only if it uses one foundry partner and one process technology for a wide array of its chips.

The post AMD may be developing APUs for TSMC’s 16nm FinFET process technology first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/amd-may-be-developing-apus-for-tsmcs-16nm-finfet-process-technology/feed/ 5
UMC skips 20nm planar process tech, leaps ahead to 16nm FinFET https://www.kitguru.net/components/anton-shilov/umc-skips-20nm-planar-process-tech-leaps-ahead-to-16nm-finfet/ https://www.kitguru.net/components/anton-shilov/umc-skips-20nm-planar-process-tech-leaps-ahead-to-16nm-finfet/#respond Thu, 30 Apr 2015 06:45:30 +0000 http://www.kitguru.net/?p=247452 United Microelectronics Corp., the world’s second largest contract maker of semiconductors, said that it would not offer 20nm planar fabrication process, but will jump directly to 16nm FinFET manufacturing technology. Skipping the node will help the company to catch up with its competitors. Although it was widely expected that UMC would offer another manufacturing technology …

The post UMC skips 20nm planar process tech, leaps ahead to 16nm FinFET first appeared on KitGuru.]]>
United Microelectronics Corp., the world’s second largest contract maker of semiconductors, said that it would not offer 20nm planar fabrication process, but will jump directly to 16nm FinFET manufacturing technology. Skipping the node will help the company to catch up with its competitors.

Although it was widely expected that UMC would offer another manufacturing technology with planar transistors after 28nm node, the firm decided to skip its 20nm process entirely and focus on development of its 16nm technology with three-dimensional FinFET transistors. The decision to leap directly to 16nm FinFET from 28nm, will help UMC to catch up with its larger rival Taiwan Semiconductor Manufacturing Co. as well as smaller competitors like GlobalFoundries and Samsung, who are either making chips using FinFET technologies or about to start production of such ICs [integrated circuits].

Not a lot is known about UMC’s upcoming 16nm manufacturing process. UMC has licensed numerous fabrication technologies from IBM and also has developed something in-house. In late 2012 the company announced that its 14nm FinFET process would be based on a modular architecture that uses 14nm FinFET devices combined with 20nm process’s back-end-of-line (BEOL) interconnect flow. It is possible that the company’s transistors and other elements ended up larger than UMC had originally anticipated and it renamed the process to “16nm”. It is also possible that UMC has developed a “true” 16nm FinFET technology with native BEOL interconnect flow with technological help from IBM.

semiconductor_umc_wafers

“We have good progress and we met the high-performance device targets,” said Po Wen Yen, chief executive officer of UMC, during the company’s conference call with investors and financial analysts. “Customers’ test chip verification is in progress.”

The company hopes to start volume production of chips using its 16nm FinFET process tech by the end of the year.

Discuss on our Facebook page, HERE.

KitGuru Says: UMC has become significantly more aggressive in the recent quarters than it used to be in the past. Still, it remains to be seen whether the company’s progress is conditioned by its collaboration with IBM, or by increased research and development efforts.

The post UMC skips 20nm planar process tech, leaps ahead to 16nm FinFET first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/umc-skips-20nm-planar-process-tech-leaps-ahead-to-16nm-finfet/feed/ 0
TSMC: 16nm yields are approaching mature levels https://www.kitguru.net/components/anton-shilov/tsmc-16nm-yields-are-approaching-mature-levels/ https://www.kitguru.net/components/anton-shilov/tsmc-16nm-yields-are-approaching-mature-levels/#respond Tue, 14 Apr 2015 10:33:20 +0000 http://www.kitguru.net/?p=244977 Although Taiwan Semiconductor Manufacturing Co. has delayed mass production of chips using its 16nm fabrication processes, this did not happen only because of low yields. According to the company, 16nm yields at TSMC are approaching mature levels. This year TSMC will offer two 16nm process technologies for clients: 16nm FinFET (CLN16FF) and 16nm FinFET+ (CLN16FF+). …

The post TSMC: 16nm yields are approaching mature levels first appeared on KitGuru.]]>
Although Taiwan Semiconductor Manufacturing Co. has delayed mass production of chips using its 16nm fabrication processes, this did not happen only because of low yields. According to the company, 16nm yields at TSMC are approaching mature levels.

This year TSMC will offer two 16nm process technologies for clients: 16nm FinFET (CLN16FF) and 16nm FinFET+ (CLN16FF+). Both technologies use FinFET transistors, but rely on the back-end-of-line (BEOL) interconnect flow of the company’s 20nm SOC (CLN20SOC) fabrication process. Usage of FinFET transistors allows to increase clock-rate potential of chips by up to 40 per cent at the same power over chips made using 20nm technology.

TSMC claims that 16nm FinFET+ provides up to 15 per cent performance improvement over the 16nm FinFET at the same level of power consumption. At the same frequency, integrated circuits produced using 16nm FinFET+ are projected to consume 30 per cent less power compared to the same chips made using 16nm FinFET.

tsmc_semiconductor_fab14_production_1

TSMC originally planned to start making 16nm chips in early 2015, but had to postpone the beginning of volume production due to undisclosed reasons. While TSMC’s motives to delay mass production were unclear, it does not look like the company has major problems with yields. According to Y.J. Mii, vice president of R&D at TSMC, CLN16FF+ yield is already approaching CLN20SoC yield (which is mature enough to use for commercial products), according to a Cadence blog post. The VP reportedly said that the 16FF+ provided better maturity at risk production than any previous TSMC process.

TSMC has received over 12 CLN16FF+ tape outs so far and expects more than 50 product tape outs this year. High-volume production will begin in the third quarter, with meaningful revenue contribution starting in the Q4 2015.

Discuss on our Facebook page, HERE.

KitGuru Says: While it looks like TSMC’s 16nm yields are reasonably high, keep in mind that the company’s 16nm chips have the same die sizes as ICs made using 20nm process tech. Since TSMC naturally charges extra for FinFET transistors as well as extra performance and lower power, per-gate costs of products made using 16nm FinFET technologies should be pretty high. As a result, to make those chips economically feasible, TSMC’s clients require generally very high yields.

The post TSMC: 16nm yields are approaching mature levels first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/tsmc-16nm-yields-are-approaching-mature-levels/feed/ 0
TSMC looking at new materials, transistor structures for sub-10nm tech https://www.kitguru.net/components/anton-shilov/tsmc-looking-at-advanced-materials-new-transistor-technologies-for-sub-10nm-tech/ https://www.kitguru.net/components/anton-shilov/tsmc-looking-at-advanced-materials-new-transistor-technologies-for-sub-10nm-tech/#respond Sat, 04 Apr 2015 18:58:29 +0000 http://www.kitguru.net/?p=243736 Taiwan Semiconductor Manufacturing Co. this week confirmed that it has a team of engineers working on a manufacturing process that will succeed 10nm fabrication technology towards the end of the decade. While the company did not reveal details about its sub-10nm process, it said that it considers various options for it, including new lithography tools, …

The post TSMC looking at new materials, transistor structures for sub-10nm tech first appeared on KitGuru.]]>
Taiwan Semiconductor Manufacturing Co. this week confirmed that it has a team of engineers working on a manufacturing process that will succeed 10nm fabrication technology towards the end of the decade. While the company did not reveal details about its sub-10nm process, it said that it considers various options for it, including new lithography tools, transistor structures and materials.

“We are working on future platform technology development,” said Suk Lee, senior director of design infrastructure marketing division of TSMC in an interview with Cadence. “We have a team working on the next generation after 10nm. Those technologies are going to be offered in the 2017 to 2019 period. We do not anticipate Moore's Law is going to slow down anytime soon.”

semiconductor_wafer_cadence_4645_TSMC-Fab3inr008_78-sized

Back in February Intel Corp. and Samsung Electronics confirmed that their research and development (R&D) teams are already working on 7nm manufacturing technologies that will be used to make highly-sophisticated computer chips. Intel confirmed that it was looking at all-new materials for its 7nm production process. Recently rumours emerged that the chipmaker is working with Honeywell, a multinational conglomerate company with a lot of expertise in chemistry and high tech, on the materials for its 7nm manufacturing tech. Samsung Electronics said that it was considering switching to a new transistor structure from FinFET/tri-gate. According to the company, gate-all-around FETs will be viable at 7nm and beyond. Samsung also implied that it would look at new materials for its 7nm fabrication process as well.

TSMC constantly makes updates regarding progress of its preparations to manufacturing using 10nm fabrication process and makes no secret that its process technology options are rather flexible. For example, TSMC is working with ASML, a leading maker of semiconductor manufacturing equipment, on a mid-node insertion of extreme ultraviolet lithography (EUV) at the 10nm logic node, which is expected to happen in late 2016. However, the company has always been tight-lipped on its processes beyond 10nm. In fact, the company even does not want to talk about exact feature size of the process.

“I do not want to make any comment on a specific technology node number,” said Mr. Lee.

But while we do not know what is exactly TSMC developing, the director of design infrastructure marketing division of TSMC revealed that the company is – just like its rivals – considering new materials and new transistor structures for its sub-10nm fabrication process.

“We are looking at advanced materials, different kinds of transistor technologies,” said Mr. Lee.

tsmc_semiconductor_fab14_production_4

In fact, keeping in mind that TSMC definitely wants to use EUV lithography at some point, it is more concerned about tools rather than transistor structures or materials. Obviously, the company is working in different directions and at some point will make decisions regarding exact materials and transistor structures for its sub-10nm process tech. However, the EUV-related challenges preserve the company from making any announcements today.

“But the immediate next big challenge has to do with continuing the ability to do patterning based on existing light sources,” stated the high-ranking TSMC rep. “At 10nm, no one is dependent on EUV. We have set up our 10nm technology so that when and if EUV comes on line we can take advantage of it. We continue to work with ASML on EUV tools. The technology itself is very complex. We need to get to a point where there's sufficient wattage and uptime so there are significant wafer volumes.”

Discuss on our Facebook page, HERE.

KitGuru Says: 7nm process technology will not be used for commercial chips until at least 2019. However, three leading semiconductor companies – Intel, Samsung and TSMC – are already developing manufacturing processes to produce chips for your 2020 smartphone, tablet and PC.

The post TSMC looking at new materials, transistor structures for sub-10nm tech first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/tsmc-looking-at-advanced-materials-new-transistor-technologies-for-sub-10nm-tech/feed/ 0
Nvidia starts to use Samsung as a chip manufacturing partner – company https://www.kitguru.net/components/anton-shilov/nvidia-starts-to-use-samsung-as-a-chip-manufacturing-partner-company/ https://www.kitguru.net/components/anton-shilov/nvidia-starts-to-use-samsung-as-a-chip-manufacturing-partner-company/#comments Fri, 20 Mar 2015 14:01:08 +0000 http://www.kitguru.net/?p=241183 Nvidia Corp. and Samsung Electronics have quietly signed an agreement under which the latter will produce computer chips for the former. At present, it is unknown whether Samsung will manufacture graphics processors or application processors for the chip designer, but it is obvious that Taiwan Semiconductor Manufacturing Co. is no longer exclusive production partner of …

The post Nvidia starts to use Samsung as a chip manufacturing partner – company first appeared on KitGuru.]]>
Nvidia Corp. and Samsung Electronics have quietly signed an agreement under which the latter will produce computer chips for the former. At present, it is unknown whether Samsung will manufacture graphics processors or application processors for the chip designer, but it is obvious that Taiwan Semiconductor Manufacturing Co. is no longer exclusive production partner of Nvidia.

“We utilize industry-leading suppliers, such as Taiwan Semiconductor Manufacturing Company Limited and Samsung Electronics Co. Ltd, to produce our semiconductor wafers,” a statement in the company’s latest 10K filing reads. Previously, similar filings only said that Nvidia uses manufacturing capacities of TSMC.

Rumours that Samsung will produce chips for Nvidia have been floating around for a number of months, but Nvidia Corp. has been downplaying them.

nvidia_tegra_x1_cut

Right now Nvidia’s lineup of products includes graphics processing units (GPUs) produced using TSMC’s 28nm fabrication process, mobile application processors made using low-power 28nm manufacturing technology as well as system-on-chips for automotive applications manufactured at TSMC’s 20nm node. Going forward Nvidia plans to make GPUs (that belong to the “Pascal” generation) and future SoCs using TSMC’s 16nm FinFET and 16nm FinFET+ process technologies, according to publicly available roadmaps.

It is unknown which processors will Samsung produce for Nvidia. Fabrication technologies at different foundries are dissimilar; hence, Samsung cannot start production of the same chips as TSMC immediately. Nvidia will have to redesign its products for Samsung’s manufacturing technologies using different libraries of elements and design tools.

samsung_semiconductor_foundry_chip_production_1

Many fabless chip developers nowadays are concerned about yields of processors that will be made using 14nm FinFET and 16nm FinFET manufacturing technologies at Samsung/GlobalFoundries and TSMC, respectively. In addition, they are concerned about availability of manufacturing capacities at leading foundries. Therefore, fabless designers plan to diversify FinFET chip suppliers, something that is expected to bring a lot of benefits to Samsung and GlobalFoundries.

Since TSMC will be late with volume production of chips using 16nm FinFET process technologies, for many makers it makes a great sense to contract Samsung, who is already making semiconductors using 14nm FinFET manufacturing technology.

NVIDIA_GeForce_GTX_900series_KeyVisual_HD_003

One analyst believes that Nvidia plans to use its contract with Samsung’s not only to get access to leading-edge fabrication processes and to ensure high-volume supply of chips, but also in order to cut-down its costs.

“We believe that Nvidia has a second source for foundry wafers in Samsung, outside of TSMC,” said Doug Freedman, an analyst with RBC Capital Markets, in a note to clients, reports SemiconductorEngineering. “While it may be a small part of the wafer supply today, it should create a tailwind for [gross margin percentage] going forward as we believe that non-TSMC wafers can be purchased at as much as 10% below present cost levels.”

No matter what Nvidia’s goals with Samsung’s foundry are, the partnership between the two companies represents a major win for Samsung’s contract manufacturing business unit. There are also rumours that Qualcomm might start using Samsung's chip production services this year. This is also a wake-up call for TSMC as its major customers are now looking at other foundries.

Discuss on our Facebook page, HERE.

KitGuru Says: It is rather ironic that as Nvidia and Samsung collaborate for chip manufacturing, the two companies are fighting in courts and accuse each other of patent infringements.

The post Nvidia starts to use Samsung as a chip manufacturing partner – company first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/nvidia-starts-to-use-samsung-as-a-chip-manufacturing-partner-company/feed/ 5
Nvidia: TSMC will remain a ‘very important’ foundry partner https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-tsmc-will-remain-a-very-important-foundry-partner/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-tsmc-will-remain-a-very-important-foundry-partner/#comments Thu, 12 Feb 2015 20:59:44 +0000 http://www.kitguru.net/?p=235446 After IBM failed to produce enough code-named “NV40” graphics processing units for Nvidia Corp. back in 2004, the latter decided to make Taiwan Semiconductor Manufacturing Co. its strategic manufacturing partner. Since then, virtually all of Nvidia GPUs have been produced by TSMC with one or two exceptions. Apparently, Nvidia wants to keep it that way …

The post Nvidia: TSMC will remain a ‘very important’ foundry partner first appeared on KitGuru.]]>
After IBM failed to produce enough code-named “NV40” graphics processing units for Nvidia Corp. back in 2004, the latter decided to make Taiwan Semiconductor Manufacturing Co. its strategic manufacturing partner. Since then, virtually all of Nvidia GPUs have been produced by TSMC with one or two exceptions. Apparently, Nvidia wants to keep it that way going forward.

Recent market rumours suggested that Nvidia plans to use manufacturing capacities of Samsung Electronics or GlobalFoundries to produce some of its multimedia or graphics processors in the future. When asked about this during a conference call with investors and financial analysts this week, Jen-Hsun Huang, chief executive officer of Nvidia, neither confirmed nor denied the plan, but he clearly stated that TSMC will remain the strategic manufacturing partner of the company.

“We always look at all foundries, and TSMC remains our most strategic [partner],” said Mr. Huang. “They are going to continue to be a very important partner for us for the foreseeable future.”

NVIDIA_GeForce_GTX_900series_KeyVisual_HD_003

Recently TSMC confirmed that it would only start to produce chips using 16nm FinFET process technology in the third quarter of the year, considerably later than Samsung Electronics and even GlobalFoundries. Moreover, given than many fabless chip designers are not confident of manufacturing capacities and yields of products made using various FinFET manufacturing processes, numerous developers create chips for different contract makers of semiconductors so to ensure stable supply.

During the conference call Nvidia revealed nothing new about its future products and fabrication processes it is going to use, but said that test chips produced using TSMC’s 16nm FinFET manufacturing technology met its criteria.

“TSMC is a fabulous supplier, […] their FinFET technology is excellent,” said the CEO of Nvidia. “[We have been] working with TSMC on FinFET now for a couple of years, and so we have quite a bit of confidence in their ability to deliver amazing FinFET transistors.”

samsung_semiconductor_foundry_chip_production_1

Nvidia is expected to release its first products made using 16nm FinFET process technology in 2016. The company’s code-named “Pascal” GPUs will be the first to be made using the new manufacturing tech.

Since prices of high-volume manufacturing using leading-edge technologies have been increasing in the recent years, Nvidia could use its potential partnership with Samsung Electronics or GlobalFoundries to negotiate better pricing with TSMC. If that is the case, then actual volumes that the two foundries are set make for Nvidia are not going to be high.

Discuss on our Facebook page, HERE.

KitGuru Says: Keeping in mind that the market of discrete graphics processing units is shrinking, it remains to be seen whether Nvidia now truly needs multiple manufacturing partners to ensure low pricing and stable supply…

The post Nvidia: TSMC will remain a ‘very important’ foundry partner first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-tsmc-will-remain-a-very-important-foundry-partner/feed/ 3
Intel: Our lead over the rest of the industry is extending https://www.kitguru.net/components/cpu/anton-shilov/intel-our-lead-over-the-rest-of-the-industry-is-extending/ https://www.kitguru.net/components/cpu/anton-shilov/intel-our-lead-over-the-rest-of-the-industry-is-extending/#respond Tue, 10 Feb 2015 23:56:07 +0000 http://www.kitguru.net/?p=235043 Intel Corp. had to delay its code-named “Broadwell” microprocessors by about a year because of yield issues with its 14nm manufacturing technology. The move has affected plans of PC makers and will have an impact on PC business this year. However, Intel’s dominant positions will hardly be threatened since the company is still ahead of …

The post Intel: Our lead over the rest of the industry is extending first appeared on KitGuru.]]>
Intel Corp. had to delay its code-named “Broadwell” microprocessors by about a year because of yield issues with its 14nm manufacturing technology. The move has affected plans of PC makers and will have an impact on PC business this year. However, Intel’s dominant positions will hardly be threatened since the company is still ahead of the rest of the industry.

“Our leadership over the rest of the industry is extending,” said Stacy Smith, chief financial officer of Intel, in an interview with Business Insider. “We are not delayed relative to the industry. We are actually ahead of the industry.”

While Intel delayed high-volume launch of its “Broadwell” chips made using 14nm fabrication process with FinFET transistors by about a year, it should be noted that its rivals are only starting to manufacture chips using their FinFET technologies. Moreover, 14nm and 16nm FinFET technologies of Samsung/GlobalFoundries and TSMC rely on back-end-of-line (BEOL) interconnect layer of 20nm fabrication process, which means that they are less technologically advanced than Intel’s 14nm.

intel_core_pentium_devil_s_canyon_lga1150_haswell

When it comes to performance and performance-per-watt, Intel’s chips are unrivalled nowadays. Microprocessors from Advanced Micro Devices are significantly behind Intel’s Core i5 or Core i7, whereas chips based on ARM architecture cannot run Windows and mainstream programs.

Thanks to unsurpassed performance and consistent performance improvements, Intel believes that it will not lose its current customers and companies like Apple will continue to use its chips.

“For a customer like Apple you’d have to take a big step off performance to step off our architecture,” said Mr. Smith. “That is what in essence enables us to win across different customers.”

Discuss on our Facebook page, HERE.

KitGuru Says: While the lead of Intel over the rest of the industry is extending in many cases, it should be noted that the industry will inevitably create certain headwinds for Intel in order to avoid its total dominance on different markets. One can see that on the market of servers, where Intel controls 98 per cent of the CPU market.

The post Intel: Our lead over the rest of the industry is extending first appeared on KitGuru.]]>
https://www.kitguru.net/components/cpu/anton-shilov/intel-our-lead-over-the-rest-of-the-industry-is-extending/feed/ 0
TSMC to invest $15.9 billion in expansion of manufacturing capacities https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-to-invest-15-9-billion-in-expansion-of-manufacturing-capacities/ https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-to-invest-15-9-billion-in-expansion-of-manufacturing-capacities/#respond Tue, 10 Feb 2015 20:34:40 +0000 http://www.kitguru.net/?p=235010 Taiwan Semiconductor Manufacturing Co. has announced that it will invest NT$500 billion ($15.9 billion) in expansion of its manufacturing capacities in central Taiwan. The company did not outline details of the upgrade plan, but it looks like the company is getting ready to significantly increase its abilities to produce semiconductors using leading edge process technologies. …

The post TSMC to invest $15.9 billion in expansion of manufacturing capacities first appeared on KitGuru.]]>
Taiwan Semiconductor Manufacturing Co. has announced that it will invest NT$500 billion ($15.9 billion) in expansion of its manufacturing capacities in central Taiwan. The company did not outline details of the upgrade plan, but it looks like the company is getting ready to significantly increase its abilities to produce semiconductors using leading edge process technologies.

The planned investment is in the Central Taiwan Science Park, near the city of Taichung, where TSMC’s fab 15 is located. The company expects the investment to create an additional 5000 jobs at the site that currently employs more than 3400 people, reports EETimes web-site. The investment program will last several years, with the amount in each year based on expected demand for capacity.

tsmc_fab2

“Taichung is a place where TSMC is well established in Taiwan,” said C.C. Wei, co-CEO of TSMC. “It is an important base where we not only want to develop advanced technology, continue revenue growth and reach new milestones, but more importantly, create a world-class green enterprise.”

At present the fab 15 produces chips using 28nm process technology. TSMC will make chips using 20nm and 16nm FinFET/16nm FinFET+ process technologies at fab 12 and fab 14 manufacturing facilities. The upgraded fab 15 complex will most likely manufacture chips using 10nm and even 7nm fabrication processes.

tsmc_semiconductor_fab14_production_4

Back in January TSMC announced that its capital expenditures for 2015 would be between $11.5 billion and $12 billion, an increase of around 20 per cent compared to the previous year. Given the amount of investments that TSMC plans to make in its site in Central Taiwan Science Park, it looks like the company plans to significantly expand the facility. What is unclear is whether the company has plans to significantly increase its CapEx.

Discuss on our Facebook page, HERE.

KitGuru Says: Even though TSMC faces stronger competition from GlobalFoundries and Samsung Electronics than ever before, it remains optimistic and expects its customers to continue using its services going forward.

The post TSMC to invest $15.9 billion in expansion of manufacturing capacities first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-to-invest-15-9-billion-in-expansion-of-manufacturing-capacities/feed/ 0
Samsung to produce chips for Nvidia, Qualcomm – analyst https://www.kitguru.net/components/graphic-cards/anton-shilov/samsung-to-produce-chips-for-nvidia-analyst/ https://www.kitguru.net/components/graphic-cards/anton-shilov/samsung-to-produce-chips-for-nvidia-analyst/#respond Wed, 04 Feb 2015 23:59:21 +0000 http://www.kitguru.net/?p=234052 Samsung Electronics is expected to significantly improve its contract semiconductor manufacturing business this year thanks to early start of production using 14nm FinFET fabrication process. While Apple, the largest customer of Samsung, now orders chips not only to Samsung, but to TSMC as well, it continues to work with the conglomerate. Moreover, due to concerns …

The post Samsung to produce chips for Nvidia, Qualcomm – analyst first appeared on KitGuru.]]>
Samsung Electronics is expected to significantly improve its contract semiconductor manufacturing business this year thanks to early start of production using 14nm FinFET fabrication process. While Apple, the largest customer of Samsung, now orders chips not only to Samsung, but to TSMC as well, it continues to work with the conglomerate. Moreover, due to concerns about capacities, a number of customers now plan to make chips at both TSMC and Samsung.

Nvidia Corp., a long-time customer of Taiwan Semiconductor Manufacturing Co., as well as Qualcomm and some other fabless semiconductor developers, plan to order chip manufacturing to Samsung Electronics this year. Many chip designers are concerned about yields of ICs that will be made using 14nm FinFET and 16nm FinFET manufacturing technologies. In addition, they are concerned about availability of manufacturing capacities. As a result, fabless companies plan to diversify FinFET chip suppliers, something that is expected to bring a lot of benefits to Samsung and GlobalFoundries.

“Global AP makers are likely to focus on lowering the cost of production this year in order to offset a decline in AP prices,” said Park Yu-ak, an analyst at Meritz Securities, in an interview with BusinessKorea. “Samsung's system semiconductor business is going to mass produce and supply chips to Apple, Qualcomm, and Nvidia starting in the second quarter of this year. As a result, the Korean tech giant is projected to exhibit stellar performance.”

It is not completely clear which chips from Nvidia will be made by Samsung, but it is an alarming news that a loyal TSMC customer wants Samsung to produce a product or two.

samsung_semiconductor_foundry_chip_production_4

Recently Nvidia introduced its Tegra X1 system-on-chip that will be made using 20nm fabrication process at TSMC.

Since Nvidia’s next-generation graphics processing units that are based on the code-named “Pascal” architecture are due only in 2016, it is unlikely that Samsung will start production of Nvidia’s new GPUs in the second quarter of 2015. It is possible that the company will shrink certain versions of its current-generation graphics chips to 14nm FinFET process to cut their power consumption, decrease manufacturing cost and/or improve performance. However, it is more likely that Samsung will make certain new Tegra application processors for Nvidia.

Nvidia, Samsung, Qualcomm and TSMC did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: If the information is correct, then this will be the first time in more than ten years when Nvidia uses a foundry other than TSMC.

The post Samsung to produce chips for Nvidia, Qualcomm – analyst first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/samsung-to-produce-chips-for-nvidia-analyst/feed/ 0
ARM introduces new-generation Cortex-A72, second-gen 64-bit core https://www.kitguru.net/components/cpu/anton-shilov/arm-introduces-new-generation-cortex-a72-second-gen-64-bit-core/ https://www.kitguru.net/components/cpu/anton-shilov/arm-introduces-new-generation-cortex-a72-second-gen-64-bit-core/#comments Tue, 03 Feb 2015 23:40:58 +0000 http://www.kitguru.net/?p=233802 ARM Holdings, a leading developer of mobile processors and solutions, on Tuesday introduced its second-generation high-performance 64-bit core based on the ARMv8-A architecture. While ARM has not revealed a lot of details about its new core, it said that it significantly increases performance compared to existing offerings and is the highest-performing general-purpose processing core that …

The post ARM introduces new-generation Cortex-A72, second-gen 64-bit core first appeared on KitGuru.]]>
ARM Holdings, a leading developer of mobile processors and solutions, on Tuesday introduced its second-generation high-performance 64-bit core based on the ARMv8-A architecture. While ARM has not revealed a lot of details about its new core, it said that it significantly increases performance compared to existing offerings and is the highest-performing general-purpose processing core that can be licensed at present.

The ARM Cortex-A72 is based on ARMv8-A architecture and is designed to run at clock-rates of up to 2.50GHz and beyond when implemented using TSMC’s 16nm FinFET+ process technology. The Cortex-A72 core can be integrated into quad-core clusters for modern smartphone or server processors. The A72-based system-on-chips will support all modern ARM technologies, including TrustZone security, NEON advanced SIMD extensions, VFPv4 floating point unit, virtualization as well as backwards compatibility with ARMv7 apps.

arm_Cortex-A72-chip-diagram-

According to ARM, the Cortex-A72 is expected to deliver roughly 80 – 90 per cent higher performance than the Cortex-A57 while consuming about the same amount of power, which means that the A72 has higher instructions per clock (IPC) rate and generally more efficient architecture. The company also claims that its new core is also 75 per cent more power efficient than the Cortex-A15 at target process technology, which is not a big surprise.

arm_cortex_72_performance

The ARM Cortex-A72 is the new “big” core from ARM that could be installed into Big.Little configurations with ARM Cortex-A53 in order to wed maximum performance and low power consumption for competitive mobile system-on-chips.

Since the Cortex-A72 is based on the ARMv8-A, not the ARMv8.1-A architecture as expected previously, software makers will not have to significantly tune their applications to take advantage of the A72’s performance improvements.

“Our new premium mobile experience IP suite with the Cortex-A72 processor delivers a decisive step forward from the compelling user experiences provided by this year's Cortex-A57 based devices,” said Pete Hutton, executive vice president and president, products groups, ARM. “For multiple generations, together with our partners, we have delivered the leading-edge of the premium mobile experience. Building on this, in 2016 the ARM ecosystem will deliver even slimmer, lighter, more immersive mobile devices that serve as your primary and only compute platform.”

More than ten partners, including HiSilicon, MediaTek and Rockchip, have already licensed the Cortex-A72 processor. The first ARM Cortex-A72 SoCs are expected to emerge in 2016.

Discuss on our Facebook page, HERE.

KitGuru Says: ARM is slowly, but surely increasing performance of it cores, which is particularly important for its server business. It should be noted, though, that the more important improvements will become available along with chips powered by the ARMv8.1-A architecture.

The post ARM introduces new-generation Cortex-A72, second-gen 64-bit core first appeared on KitGuru.]]>
https://www.kitguru.net/components/cpu/anton-shilov/arm-introduces-new-generation-cortex-a72-second-gen-64-bit-core/feed/ 3
AMD develops ‘Arctic Islands’ family of graphics processing units https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-develops-arctic-islands-family-of-graphics-processing-units/ https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-develops-arctic-islands-family-of-graphics-processing-units/#comments Tue, 27 Jan 2015 23:59:13 +0000 http://www.kitguru.net/?p=232810 Advanced Micro Devices yet has to reveal its new families of graphics processing units this year, but already now the code-name of AMD’s 2016 graphics products lineup is known. Apparently, the company will continue to use names of islands to call its next-generation chips. AMD’s next-generation family of Radeon GPUs will be called “Arctic Islands”, …

The post AMD develops ‘Arctic Islands’ family of graphics processing units first appeared on KitGuru.]]>
Advanced Micro Devices yet has to reveal its new families of graphics processing units this year, but already now the code-name of AMD’s 2016 graphics products lineup is known. Apparently, the company will continue to use names of islands to call its next-generation chips.

AMD’s next-generation family of Radeon GPUs will be called “Arctic Islands”, reports SweClockers. Therefore, expect code-names like Greenland, Svalbard, Ellesmere, Wrangel and others. At present nothing particular is known about the new family, except of the fact that it will traditionally cover every market segment from top to bottom. It is likely that the new graphics chips will be made using 14nm or 16nm FinFET process technologies.

AMD’s product families like “Southern Islands”, “Sea Islands” or “Volcanic Islands” may, or may not, be based on the same architecture. Traditionally, families contain chips that may belong to different generations and may have different feature-sets.

amd_radeon_artwork_angle_new

It is anticipated that AMD will reveal its all-new family of graphics processing units next year. The lineup will be based on a post-GCN architecture and will therefore bring-in new levels of performance and new functionality. Unfortunately, it is unknown whether the new GPUs will carry “Artic Islands” code-names.

It is expected that the “Arctic Islands” products will be sold under AMD Radeon R* 400-series brand names.

AMD did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: A good news is that AMD seems to have a roadmap for 2016 and will have products to sell next year. Unfortunately, we have no idea how competitive will those products be and what levels of performance should we expect from them. In many cases AMD’s new product families are not all-new, but contain loads of previous-generation offerings.

The post AMD develops ‘Arctic Islands’ family of graphics processing units first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-develops-arctic-islands-family-of-graphics-processing-units/feed/ 2
AMD: We are actively designing 14nm FinFET chips https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-we-are-actively-designing-14nm-finfet-chips/ https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-we-are-actively-designing-14nm-finfet-chips/#comments Wed, 21 Jan 2015 20:59:13 +0000 http://www.kitguru.net/?p=231884 Advanced Micro Devices on Tuesday confirmed for the first time that it is actively designing several chips that will be made using 14nm FinFET process technology. The announcement from AMD indicates that the company will manufacture its new chips either at GlobalFoundries or Samsung Electronics, not at Taiwan Semiconductor Manufacturing Co., which is a huge …

The post AMD: We are actively designing 14nm FinFET chips first appeared on KitGuru.]]>
Advanced Micro Devices on Tuesday confirmed for the first time that it is actively designing several chips that will be made using 14nm FinFET process technology. The announcement from AMD indicates that the company will manufacture its new chips either at GlobalFoundries or Samsung Electronics, not at Taiwan Semiconductor Manufacturing Co., which is a huge surprise.

“We are actively designing a number of products in the 14nm FinFET process technology,” said Lisa Su, chief executive officer of AMD, during a conference call with investors and financial analysts.

globalfoundries_semiconductor_wafers_300mm

TSMC, the world’s largest contract maker of semiconductors, will offer its customers 16nm FinFET and 16nm FinFET+ manufacturing technologies starting from Q3 2015. By contrast, GlobalFoundries and Samsung Electronics will use 14nm LPE (low power early) and 14nm LPP (low power plus) fabrication processes to produce chips for their clients. In fact, Samsung is already making semiconductors using 14nm LPE technology, whereas GlobalFoundries expects to start making 14nm chips sometimes in the first half of the year.

AMD did not reveal which of its chips will be made using 14nm FinFET fabrication process and which of the two technologies will be used. However, it is logical to expect the company to utilise various 14nm technologies for all types of its products, including accelerated processing units, central processing units, graphics processing units as well as semi-custom chips for various applications.
amd_apu_beema_mullins_puma_jaguar_x86_fusion

Separately, AMD confirmed that it would reveal more details about its long-term roamdaps in the coming months at the financial analyst day.

“Relative to timing on FinFET, we will be talking more about our long-term roadmaps at our financial analyst day,” said Ms. Su.

Discuss on our Facebook page, HERE.

KitGuru Says: If AMD plans to use 14nm FinFET instead of 16nm FinFET, this will mean that it will cut its orders to TSMC. This could be an alarming sign for TSMC since AMD is one of the largest customers of the company.

The post AMD: We are actively designing 14nm FinFET chips first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/amd-we-are-actively-designing-14nm-finfet-chips/feed/ 8
TSMC unveils third 16nm process for ultra-low-power, high-performance devices https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-unveils-third-16nm-process-for-ultra-low-power-high-performance-devices/ https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-unveils-third-16nm-process-for-ultra-low-power-high-performance-devices/#respond Fri, 16 Jan 2015 08:10:31 +0000 http://www.kitguru.net/?p=230879 Taiwan Semiconductor Manufacturing Co. has confirmed that it is developing yet another version of its 16nm FinFET fabrication process. The new manufacturing technology will be used to make chips for ultra-low-power as well as high-performance applications. “We are also working on 16nm ULP technology development,” said Mark Liu, co-CEO of TSMC, during a conference call …

The post TSMC unveils third 16nm process for ultra-low-power, high-performance devices first appeared on KitGuru.]]>
Taiwan Semiconductor Manufacturing Co. has confirmed that it is developing yet another version of its 16nm FinFET fabrication process. The new manufacturing technology will be used to make chips for ultra-low-power as well as high-performance applications.

“We are also working on 16nm ULP technology development,” said Mark Liu, co-CEO of TSMC, during a conference call with investors and financial analyst. “This 16nm ULP design kits will be available in June this year. It will be suitable for both high performance and ultra-low power or ultra-low voltage, less than 0.6V applications.”

At present TSMC offers 16nm FinFET (CLN16FF) and 16nm FinFET+ (CLN16FF+) manufacturing technologies to its customers. Both provide significant performance and power saving improvements compared to TSMC’s 28nm and 20nm fabrication processes thanks to FinFET transistors. While the 16FF and the 16FF+ share the same metal backend process with 20nm SOC technology and therefore cannot reduce chip area, semiconductors made using the 16nm processes are considerably smaller than those produced using 28nm tech.

tsmc_semiconductor_fab12_1

About a year ago rumours emerged that TSMC was developing its third 16nm process technology, which was supposed to be called 16nm FinFET Turbo. It looks like the 16nm FinFET ULP technology is the 16nm FinFET Turbo as it is expected to provide certain clock-rate improvements for devices that need it. In addition, the 16nm ULP manufacturing process will help TSMC to cut-down power consumption of system-on-chips for Internet-of-Things and wearable devices.

As usually, TSMC did not unveil a lot of information regarding the 16nm FinFET ultra-low-power process technology. It is not clear when TSMC plans to start making 16nm FinFET ULP chips in volume, but it is logical to expect such chips to be ready in late 2016 or early 2017.

Discuss on our Facebook page, HERE.

KitGuru Says: Three versions of 16nm FinFET fabrication processes indicate that the node will be used for quite a long time.

The post TSMC unveils third 16nm process for ultra-low-power, high-performance devices first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/tsmc-unveils-third-16nm-process-for-ultra-low-power-high-performance-devices/feed/ 0
TSMC delays mass production using 16nm FinFET process to Q3 2015 https://www.kitguru.net/components/anton-shilov/tsmc-delays-mass-production-using-16nm-finfet-process-to-q3-2015/ https://www.kitguru.net/components/anton-shilov/tsmc-delays-mass-production-using-16nm-finfet-process-to-q3-2015/#respond Fri, 16 Jan 2015 00:00:09 +0000 http://www.kitguru.net/?p=230858 Taiwan Semiconductor Manufacturing Co. has clarified the timeframe when it intends to start mass production of chips using 16nm FinFET fabrication process. The world’s largest contract maker of semiconductors once again will delay 16nm FF volume production by a month or two to the third quarter of 2015. Back in October, 2014, TSMC already said …

The post TSMC delays mass production using 16nm FinFET process to Q3 2015 first appeared on KitGuru.]]>
Taiwan Semiconductor Manufacturing Co. has clarified the timeframe when it intends to start mass production of chips using 16nm FinFET fabrication process. The world’s largest contract maker of semiconductors once again will delay 16nm FF volume production by a month or two to the third quarter of 2015.

Back in October, 2014, TSMC already said that it would move mass production of chips using its 16nm FinFET manufacturing technology from early 2015 to Q2 or early Q3 2015. This week the company said that the aforementioned semiconductors will be made only sometimes in the third quarter of the year, which essentially means another delay, albeit, a slight one.

“We plan to mass produce 16nm FinFET [chips] in the third quarter of 2015,” said Lora Ho, senior vice president and chief financial officer of TSMC.

It will take the company around seven quarters to bring profit margins it receives from 16nm FinFET manufacturing to corporate average levels, according to the CFO. This year the company plans to produce 50 chip designs using the 16nm FinFET fabrication process.

tsmc_semiconductor_fab14_production_4

TSMC’s 16nm FinFET (CLN16FF)and 16nm FinFET+ (which TSMC officially calls 16nm FinFET Plus) process technologies rely on the back-end-of-line (BEOL) interconnect flow of the company’s 20nm SOC (CLN20SOC) fabrication process, but use FinFET transistors instead of planar transistors. This provides additional performance and/or power savings, but this does not allow to significantly shrink the size of chips compared to chips made using the 20nm SOC. The proven BEOL interconnect flow means that it should be easier for TSMC to start mass production of chips using its 16FF and 16FF+ manufacturing technologies.

It is unclear how delay of mass production using 16nm FinFET process technology affects schedules of the 16nm FinFET+.

TSMC expects the capital expenditures for 2015 to be between $11.5 billion and $12 billion. 80 per cent of the company’s CapEx will be spent on development of 10nm and thinner manufacturing technologies, tools necessary to make chips using 16nm and 20nm fabrication processes as well as on other leading-edge technologies-related things.

Discuss on our Facebook page, HERE.

KitGuru Says: The situation with FinFET process technologies in general remains blurry. It is unclear when GlobalFoundries plans to start production chips using 14nm FinFET process it licensed from Samsung.  Moreover, the volumes of 14nm FinFET chip wafers processed at Samsung are a complete mystery.

The post TSMC delays mass production using 16nm FinFET process to Q3 2015 first appeared on KitGuru.]]>
https://www.kitguru.net/components/anton-shilov/tsmc-delays-mass-production-using-16nm-finfet-process-to-q3-2015/feed/ 0
Nvidia committed to its ‘Denver’ processor, will use it in future SoCs https://www.kitguru.net/components/cpu/anton-shilov/nvidia-committed-to-its-denver-processor-will-use-it-in-future-designs/ https://www.kitguru.net/components/cpu/anton-shilov/nvidia-committed-to-its-denver-processor-will-use-it-in-future-designs/#respond Tue, 06 Jan 2015 05:10:20 +0000 http://www.kitguru.net/?p=229064 Although Nvidia Corp. decided not to use its custom ARMv8-compatible 64-bit “Denver” core inside its Tegra X1 system-on-chip, the company claims that it can use it inside other application processors. What is unclear is when and inside which application processors. “The reason we did not use Denver on Tegra X1 is that we decided to …

The post Nvidia committed to its ‘Denver’ processor, will use it in future SoCs first appeared on KitGuru.]]>
Although Nvidia Corp. decided not to use its custom ARMv8-compatible 64-bit “Denver” core inside its Tegra X1 system-on-chip, the company claims that it can use it inside other application processors. What is unclear is when and inside which application processors.

“The reason we did not use Denver on Tegra X1 is that we decided to take a ‘tick tock’ approach to launching our processors,” an official statement by Nvidia published by ArsTechnica reads. “Since TX1 is on a brand new 20nm process we decided to use off the shelf ARM cores. But Tegra K1 was on a well understood [28nm] process and hence we decided to use Denver on it. We still have Denver on our roadmap and will be using it in our future chips as processes mature.”

nvidia_tegra_chip

Nvidia first revealed details about its code-named Denver general-purpose 64-bit custom ARMv8-compatible core in early 2013. The core was expected to first emerge in Nvidia Tegra design code-named “Parker”, which was supposed to feature “Maxwell” graphics and to be made using 16nm FinFET process technology. The “Parker” was due in 2015.

nvidia_tegra_roadmap

As soon as it turned out that TSMC’s 16nm FinFET fabrication process would not be ready for mass production in the first half of 2015, Nvidia decided to incorporate “Denver” into the 64-bit version of Tegra K1 “Logan” (with “Kepler” graphics), which is made using 28nm fabrication process. The “Parker” chip was thrown out from Nvidia’s roadmap and the code-named “Erista” SoC replaced it.

nvidia_tegra_roadmap_erista_anandtech
Tegra roadmap, image by AnandTech

The “Denver” core was developed from the ground-up for the 16nm FinFET process technology. Integrating the core into a 28nm design was hardly an optimal choice given die size, power consumption and other factors. Nvidia has never explained why it needed to create 64-bit Tegra K1 chip based on the “Denver” cores. Most probably, it needed software developers to lay their hands on its custom microprocessor design in order to ensure that future apps will be fully optimized for Nvidia’s high-performance ARMv8 cores.

It is logical to expect Nvidia to incorporate “Denver” into its application processors to be made using 16nm FinFET process technology, just like it planned originally with the code-named “Parker” AP. Unfortunately, since TSMC’s 16nm FinFET manufacturing technology is at least two quarters away, it is unknown when Nvidia will be able to commercialize such chip. Depending on design readiness, this could happen only late this year at the earliest. Keeping in mind Nvidia’s typical Tegra schedules, expect the new chip to arrive only in early 2016.

Discuss on our Facebook page, HERE.

KitGuru Says: It looks that the mess TSMC and other foundries created with availability of process technologies also generated a mess with product designs. As a result, Nvidia had to redesign its “Maxwell” graphics architecture as well as to change its Tegra roadmap significantly.

The post Nvidia committed to its ‘Denver’ processor, will use it in future SoCs first appeared on KitGuru.]]>
https://www.kitguru.net/components/cpu/anton-shilov/nvidia-committed-to-its-denver-processor-will-use-it-in-future-designs/feed/ 0