GM107 | KitGuru https://www.kitguru.net KitGuru.net - Tech News | Hardware News | Hardware Reviews | IOS | Mobile | Gaming | Graphics Cards Wed, 18 Mar 2015 03:39:03 +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 GM107 | KitGuru https://www.kitguru.net 32 32 Nvidia ‘Pascal’ GPUs to offer up to 10X higher speed than ‘Maxwell’ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-next-gen-pascal-gpus-to-offer-up-to-10-times-higher-performance-than-maxwell/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-next-gen-pascal-gpus-to-offer-up-to-10-times-higher-performance-than-maxwell/#comments Wed, 18 Mar 2015 01:55:00 +0000 http://www.kitguru.net/?p=240589 At its GPU Technology Conference Nvidia Corp. revealed some details about its next-generation code-named “Pascal” architecture of graphics processing units. As expected, the new GPUs will feature numerous innovations that will provide significant benefits in performance in various types of applications. Nvidia noted three key technologies that the “Pascal” GPUs will have: mixed precision computing, support for …

The post Nvidia ‘Pascal’ GPUs to offer up to 10X higher speed than ‘Maxwell’ first appeared on KitGuru.]]>
At its GPU Technology Conference Nvidia Corp. revealed some details about its next-generation code-named “Pascal” architecture of graphics processing units. As expected, the new GPUs will feature numerous innovations that will provide significant benefits in performance in various types of applications.

Nvidia noted three key technologies that the “Pascal” GPUs will have: mixed precision computing, support for stacked multi-layer 3D memory and NVLink GPU bus. As reported previously, the “Pascal” GPUs will be made using 16nm FinFET process technology, therefore, expect increased count of stream processors, and higher clock-rates in addition to architectural enhancements. According to Nvidia, in peak cases, when all improvements come into play, “Pascal” GPUs can be 10 times faster than “Maxwell” GPUs. In typical cases expect next-gen GPUs to be around two times faster than contemporary graphics chips.
nvidia_artworklatesteyeweb_mid
Mixed-precision computing enables “Pascal” architecture-based GPUs to compute at 16-bit floating point accuracy at twice the rate of 32-bit floating point accuracy, which is useful for a number of GPGPU tasks. In certain cases GPUs could do graphics work with 16-bit FP precision without degradation of quality, but not in all of them.

Stacked multi-layer 3D memory will increase memory bandwidth available to new GPUs. Nvidia does not reveal what kind of memory it plans to use, but depending on the Pascal’s availability timeframe, it could be first-generation HBM (with up to 640GB/s bandwidth) or second-generation HBM with unprecedented capacities and north from 1TB/s bandwidth. Extreme memory bandwidth brings benefits to virtually all types of applications, including video games. Graphics cards featuring “Pascal” GPUs and second-gen HBM will carry up to 32GB of memory, according to Nvidia.

NVLink is an energy-efficient, high-bandwidth bus that will be used for communications between Nvidia GPUs and IBM Power processors in supercomputers as well as for multi-GPU communications. Initial NVLink data-rates will be “at least” 80GB/s, but eventually NVLink’s bandwidth will be increased to 192GB/s or even higher.

Nvidia plans to release its first “Pascal” GPUs in 2016.

Discuss on our Facebook page, HERE.

KitGuru Says: While “Pascal” GPUs will not be 10 times faster than “Maxwell” graphics processors in all types of applications, expect the upcoming GPUs to offer significantly higher performance than predecessors due to increased amount of execution units, higher clock-rates, massively increased memory bandwidth and architectural enhancements. The majority of innovations will be enabled by thinner manufacturing process.

The post Nvidia ‘Pascal’ GPUs to offer up to 10X higher speed than ‘Maxwell’ first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-next-gen-pascal-gpus-to-offer-up-to-10-times-higher-performance-than-maxwell/feed/ 26
Nvidia readies GeForce GTX 950M and 960M graphics adapters https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-readies-geforce-gtx-950m-and-960m-graphics-adapters/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-readies-geforce-gtx-950m-and-960m-graphics-adapters/#comments Tue, 20 Jan 2015 08:14:20 +0000 http://www.kitguru.net/?p=231352 Nvidia Corp. has added support for three new GeForce mobile GPU products into its drivers. While the new GeForce GTX will belong to the 900-series, they will be powered by the GM107 graphics processing unit introduced a year ago. Laptopvideo2Go web-site reports Nvidia has added support for three new products – GeForce 940M, GeForce GTX …

The post Nvidia readies GeForce GTX 950M and 960M graphics adapters first appeared on KitGuru.]]>
Nvidia Corp. has added support for three new GeForce mobile GPU products into its drivers. While the new GeForce GTX will belong to the 900-series, they will be powered by the GM107 graphics processing unit introduced a year ago.

Laptopvideo2Go web-site reports Nvidia has added support for three new products – GeForce 940M, GeForce GTX 950M and GeForce GTX 960M – into its drivers. The GTX 950M and the GTX 960M have 139A and 139B device IDs and are based on the GM107 graphics processing unit. The GeForce 940M has 1347 device ID and is powered by an unknown graphics processor.

nvidia_artwork

It is interesting to note that so far Nvidia has not introduced any mobile solutions based on its code-named GM206 graphics processor. The GeForce GTX 965M is powered by the GM204 graphics chip with half of execution units disabled. While it is logical for Nvidia to use the GM206 for its GeForce GTX 960M, the company will use its GM107 chip for this solution.

Nvidia did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: It is unclear why Nvidia needs products like the GeForce 940M. The graphics adapter hardly provides decent performance in video games and will hardly become popular among gamers, which means that its sales will be negligible.

The post Nvidia readies GeForce GTX 950M and 960M graphics adapters first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-readies-geforce-gtx-950m-and-960m-graphics-adapters/feed/ 1
Nvidia is ‘excited’ about ‘Pascal’ and next-gen process technologies https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-we-are-excited-about-pascal-and-next-generation-process-technologies/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-we-are-excited-about-pascal-and-next-generation-process-technologies/#comments Sat, 08 Nov 2014 13:09:21 +0000 http://www.kitguru.net/?p=221054 Jen-Hsen Huang, chief executive officer of Nvidia Corp. said this week that he is excited about the company’s next-generation graphics processing units code-named “Pascal” as well as next-generation process technologies. Nevertheless, while the CEO of Nvidia is confident of the company’s roadmap, he notes that current-gen “Maxwell” family of GPUs is only beginning its journey. …

The post Nvidia is ‘excited’ about ‘Pascal’ and next-gen process technologies first appeared on KitGuru.]]>
Jen-Hsen Huang, chief executive officer of Nvidia Corp. said this week that he is excited about the company’s next-generation graphics processing units code-named “Pascal” as well as next-generation process technologies. Nevertheless, while the CEO of Nvidia is confident of the company’s roadmap, he notes that current-gen “Maxwell” family of GPUs is only beginning its journey.

“We have got lot of great surprises for you guys and I am excited about our next generation GPUs,” said Jen-Hsun Huang during quarterly conference call with investors and financial analysts. “But right now we are enjoying ramping Maxwell. This is a brand new product cycle.”

NVIDIA_GeForce_GTX_900series_KeyVisual_HD_003

Right now Maxwell mainly addresses the market of consumer gaming PCs with GeForce graphics cards. Nvidia plans to introduce professional-grade Quadro graphics cards as well as Tesla accelerators for high-performance computing applications sometime in 2015.

At present there are only two graphics processing units – GM107 and GM204 – based on the Maxwell architecture. Nvidia is expected to unveil two more chips based on the latest graphics processing technology, which will take a quarter or two. It is believed that one of the forthcoming Maxwell GPUs – code-named GM200 – will address the markets of high-performance computing, professional graphics as well as ultra-high-end gaming PCs.

Nvidia’s next-generation graphics processors are code-named “Pascal”. Based on the company’s roadmap that it demonstrated back in March, Pascal GPUs are due sometime in 2016. Next-generation graphics chips from Nvidia will support stacked high-bandwidth dynamic random access memory (DRAM) (SK Hynix's high-bandwidth memory (HBM) or Micron's hybrid memory cube (HMC)), unified memory addressing for CPU and GPU, NVLink interconnection for high-performance computing platforms as well as new graphics, compute and multimedia features that will be a part of DirectX 12, OpenGL 5.0 and other forthcoming application programming interfaces.

nvidia_pascal_module

In 2016 displays with ultra-high-definition (UHD) resolutions like 4K (3840*2160, 4096*2160) or 5K (5120*2160) will get much more popular than they are today. Therefore, dramatically improved graphics processing horsepower of Pascal GPUs as well as extreme bandwidth provided by stacked HMC or HBM DRAM devices (we are talking about 1TB/s – 2TB/s bandwidth here) will be appreciated by the market.

Given the availability timeframe of the “Pascal” family of graphics processors, it is very likely that they will be manufactured using 16nm FinFET+ process technology at Taiwan Semiconductor Manufacturing Co. Although Jen-Hsun Huang has not confirmed anything about Pascal, he did indicate that he is happy with the forthcoming fabrication processes.

“We are excited about the next generation FinFET [manufacturing technologies],” said Mr. Huang. “I can tell you that for the next couple of nodes, I feel pretty good about [them].”

Discuss on our Facebook page, HERE.

KitGuru Says: While Nvidia’s Pascal architecture looks like another major step in the evolution of graphics processors in general, it is at least 1.5 years away. Therefore, from an end-user point of view it is more interesting to know, what the GM200 is and how fast it is. The GM200 will power Nvidia’s next-generation GeForce GTX Titan graphics cards and it is not a secret that they are performance monsters.

The post Nvidia is ‘excited’ about ‘Pascal’ and next-gen process technologies first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-we-are-excited-about-pascal-and-next-generation-process-technologies/feed/ 17
Nvidia to skip GeForce GTX 800 series, to introduce GeForce GTX 970, GTX 980 in mid-September https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-to-skip-geforce-gtx-800-series-to-introduce-geforce-gtx-970-gtx-980-in-mid-september/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-to-skip-geforce-gtx-800-series-to-introduce-geforce-gtx-970-gtx-980-in-mid-september/#comments Fri, 29 Aug 2014 03:23:26 +0000 http://www.kitguru.net/?p=209448 Although it was expected that Nvidia Corp. would release its new GeForce GTX 800-series graphics cards for desktops this September and October, the company has decided to skip the GeForce 800 family for desktop PCs and jump straight to the GeForce 900 line of products. Model numbers of modern central processing units and graphics processing units have …

The post Nvidia to skip GeForce GTX 800 series, to introduce GeForce GTX 970, GTX 980 in mid-September first appeared on KitGuru.]]>
Although it was expected that Nvidia Corp. would release its new GeForce GTX 800-series graphics cards for desktops this September and October, the company has decided to skip the GeForce 800 family for desktop PCs and jump straight to the GeForce 900 line of products.

Model numbers of modern central processing units and graphics processing units have been criticized criticized for years since in many cases they barely reflect performance and positioning of products and in many cases cannot even determine their generation. At present Nvidia offers high-end GeForce GTX 700-series of products for desktop computers. At the same time, the company sells GeForce GTX 800M family of products for notebooks. What is rather confusing is that both lineups – the GTX 700 and the GTX 800M – are based on similar chips powered by the Kepler architecture.

In a bid to avoid confusion with its upcoming families of graphics processing units for desktop and mobile computers, Nvidia decided to sell its products powered by the second-generation Maxwell GPUs code-named GM200, GM204 and other under the GeForce 900 moniker, reports VideoCardz.

nvidia_artwork

As a result, Nvidia is now expected to formally introduce the GeForce GTX 970 and the GeForce GTX 980 graphics cards on the 19th of September, 2014. The GeForce GTX 960 is projected to arrive sometime in October.

Nvidia did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: While the idea to sell similar generation of chips under similar model numbers makes a lot of sense, it should be kept in mind that in order to complete the GeForce 900 and the GeForce 900M product lines, Nvidia will likely add old chips into new families and therefore will confuse end-users again.

The post Nvidia to skip GeForce GTX 800 series, to introduce GeForce GTX 970, GTX 980 in mid-September first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-to-skip-geforce-gtx-800-series-to-introduce-geforce-gtx-970-gtx-980-in-mid-september/feed/ 87
Nvidia refreshes professional lineup with new Quadro graphics cards https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-refreshes-professional-lineup-with-new-quadro-graphics-cards/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-refreshes-professional-lineup-with-new-quadro-graphics-cards/#respond Wed, 13 Aug 2014 03:20:32 +0000 http://www.kitguru.net/?p=207096 Nvidia Corp. on Tuesday refreshed the lineup of Quadro professional graphics cards. The new graphics boards are based on either Kepler or Maxwell architectures, will offer higher performance than predecessors as well as a number of new capabilities that will significantly change the way people use professional graphics cards. The new Quadro K5200, K4200, K2200, …

The post Nvidia refreshes professional lineup with new Quadro graphics cards first appeared on KitGuru.]]>
Nvidia Corp. on Tuesday refreshed the lineup of Quadro professional graphics cards. The new graphics boards are based on either Kepler or Maxwell architectures, will offer higher performance than predecessors as well as a number of new capabilities that will significantly change the way people use professional graphics cards.

The new Quadro K5200, K4200, K2200, K620 and K420 graphics cards – are based on the GK110GL, GK104GL or GM107GL graphics processing units, just like some of their processors. In addition to higher – by up to 40 per cent – performance in a number of cases, the new graphics adapters support several important functions not available on previous-generation solutions.

Quadro_Kepler_Family2

In particular, the new Quadro graphics boards allow to:

  • Interact with data sets or designs up to twice the size handled by previous generations thanks to larger memory buffers;
  • Remotely interact with graphics applications from a Quadro-based workstation from essentially any device, including PCs, Macs and tablets;
  • Switch seamlessly from local GPU rendering to cloud-based offerings using Nvidia Iray technology.

nvidia_quadro_2014_specifications

“The next generation of Quadro GPUs not only dramatically increases graphics and compute performance to handle huge data sets. It extends the concept of visual computing from a graphics card in a workstation to a connected environment,” said Jeff Brown, vice president of professional visualisation at Nvidia. “The new Quadro line-up lets users interact with their designs or data locally on a workstation, remotely on a mobile device or in tandem with cloud-based services.”

The new Quadro GPU family will ship beginning this fall from Nvidia workstation OEMs, including HP, Dell, Lenovo and other major workstation makers; from systems integrators, including Boxx Technologies and Supermicro; and from authorised distribution partners, including PNY Technologies in North America and Europe, ELSA and Ryoyo in Japan, and Leadtek in Asia Pacific.

Discuss on our Facebook page, HERE.

KitGuru Says: The new Quadro graphics cards have gained support for modern technologies, such as remote interaction with Quadro-based workstations and switching from local GPU rendering to Iray cloud rendering. It is likely that enterprises will love the new solutions and their capabilities.

The post Nvidia refreshes professional lineup with new Quadro graphics cards first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-refreshes-professional-lineup-with-new-quadro-graphics-cards/feed/ 0
Nvidia GeForce GTX 800: what to expect from its specs? https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-geforce-gtx-800-what-to-expect-from-its-performance-and-specs/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-geforce-gtx-800-what-to-expect-from-its-performance-and-specs/#comments Fri, 08 Aug 2014 22:59:18 +0000 http://www.kitguru.net/?p=206623 Nvidia Corp. and its partners among makers of graphics cards are expected to release the next-generation GeForce GTX 800-series solutions sometimes in October (or even late September), according to multiple reports. But while the new graphics solutions are just a couple of months away, it is completely unknown what to expect from them. We decided …

The post Nvidia GeForce GTX 800: what to expect from its specs? first appeared on KitGuru.]]>
Nvidia Corp. and its partners among makers of graphics cards are expected to release the next-generation GeForce GTX 800-series solutions sometimes in October (or even late September), according to multiple reports. But while the new graphics solutions are just a couple of months away, it is completely unknown what to expect from them. We decided to analyze what we do know for sure about the GM204, the graphics processor that will power the new cards, and what kind of specifications and performance should we expect.

nvidia_artwork_iron

Facts we know for (almost) sure

While there are not a lot of reliable facts from sources close to Nvidia available at the moment, here is what we have found so far about the GM204 and what either is trustworthy or, at least, looks trustworthy:

  • Nvidia GM204 is based on the code-named Maxwell architecture;
  • Nvidia GM204 is made using 28nm process technology (given that multiple sources said that the GM204 is a 28nm GPU, it is most probably a fact);
  • Nvidia GM204 has die size of around 300mm² (we’ve seen the chip, we’ve discussed that performance GPUs from Nvidia feature about 300mm² die size);
  • Nvidia GM204 most likely 256-bit memory bus (we’ve seen a GM204-based card, it features 16 GDDR5 memory chips (with 16-bit or 32-bit interfaces), hence, given the size of the chip, 256-bit memory bus is a more likely option);

While we have no idea about exact specifications of the GM204, we do know specifications of the GM107 thanks to the official launch of the GeForce GTX 750-series GPUs earlier this year and hence we know something about the Maxwell architecture in general.

Maxwell architecture

Nvidia’s graphics processors consist of several key building blocks: main scheduler/dispatch processor (which Nvidia calls GigaThread Engine), GPCs [graphics processing clusters], SMMs [streaming multiprocessor module], ROPs [raster operating units], cache, memory controllers.

GeForce_GTX_750_Ti_Block_Diagram_FINAL

A block diagram of Nvidia's GM107 graphics processing unit.

The GM107 graphics processor based on the first-gen Maxwell architecture features one GPC unit, 16 ROPs (in two blocks,), 2MB L2 cache and two 64-bit memory controllers. The GPC of the GM107 consists of five SMMs. Each first-gen Maxwell SMM features its own PolyMorph Engine 2.0 unit (one vertex fetch, one tessellator, one viewport transform, one stream output, etc.) and is divided into four blocks with their own control logic (instruction buffer, warp scheduler, dispatch unit), a register file, two texture units (TUs) and 32 stream processors (SPs) (thus, each SMM features 128 SPs, 8 TUs, etc.).

nvidia_maxwell_smm

A block diagram of a Maxwell SMM

According to Nvidia, thanks to better programmability of each block within an SMM and higher utilization rate of each stream processors, the company increased the peak performance per stream processor by 35% when compared to the SPs in its chips built with the previous generation Kepler architecture. Therefore a Maxwell-based GPU with equal (or more or less equal) amount of stream processors with a Kepler-based GPU will perform around 35 per cent better.

Nvidia’s GM107 graphics processing unit features 640 SPs, 40 TUs, 16 ROPs, 2MB cache and a 128-bit memory controller. The chip contains 1.87 billion of transistors and has 148mm² die size.

What Nvidia needs from GM204?

Since the GM204 is a successor to the GK104 graphics processor (GeForce GTX 670, 680, 760 and 770) made using the same 28nm process technology, Nvidia cannot really expect it to outperform its predecessor by two times or something like that. Realistically, Nvidia needs to tangibly outperform the best GK104 by around 30 to 40 per cent. It is impossible that a 300mm² Maxwell chip could beat the GK110 GPU with 2880 stream processors (which die size is 551mm²), hence, it is not a target that the GM204 should achieve.

Possible GM204 configurations

Each architecture can be scaled to offer better performance or lower power consumption. Let’s try to assume how Nvidia could scale the GM107’s architecture to offer better performance with the GM204 while maintaining around 300mm² die size.

The easiest way to double performance of the GM107 is to double the amount of virtually everything inside: GPC, ROPs, cache, memory controllers. Such a chip would have 1280 stream processors, 80 texture units, 32 ROPs and a 256-bit memory bus. The GPU would never outperform the GK104 (1536 SPs, 128 TUs, 32 ROPs, etc.) significantly and in many cases will be behind the older solution. Therefore, it is unlikely that Nvidia will take this route. Moreover, such a chip would be significantly smaller than 300mm² since not all elements of the chip have to be doubled.

What Nvidia did with GK110 compared to the GK104 architecturally (at least when it comes to organization of execution units) was the increase of the amount of SMs [streaming multiprocessor] per GPC from two to three. The same approach could be used for the GM204 too. Nvidia could expand GPC to six SMMs and then double the amount of GPCs (and ROPs, cache, memory controllers) per chip. Such a GPU would feature 1536 SPs, 96 TUs, 32 ROPs as well as a 256-bit memory bus and could offer up to 35 per cent higher performance compared to the GK104 at the same clock-rate. The die size of such chip would probably be around 300mm², but since we do not know anything about exact sizes of the GM107’s elements, we cannot be 100 per cent sure.

Theoretically, Nvidia could put six blocks into each SMM (thus increasing the amount of SPs per SMM to 192), then double the amount of GPCs (along with ROPs, cache, memory controllers) per GPU. If Nvidia manages to do this, then the final chip would feature 1920 SPs, 120 TUs, 32 ROPs and a 256-bit memory controller. Would that all fit into a 300mm² die area? Possibly. However, since this fundamentally changes the architecture of the SMM, this could also affect efficiency of Maxwell architecture in general. Therefore, the expansion of GPC to seven SMMs (and doubling the amount of GPCs, etc.) sounds more realistic (the GPU would have 1792 SPs, 112 TUs, 32 ROPs, 256-bit memory bus).

Final words

While chip designers can do many unexpected things and sometimes even wonders, they cannot overrule the laws of physics. Engineers will not be able to squeeze 3200 stream processors into the GM204 that is made using 28nm process technology and has die size of around 300mm². Therefore, the most logical configurations of the fully-fledged GM204 (the GeForce GTX 880) are as follows:

  • 1536 SPs, 96 TUs, 32 ROPs, 256-bit memory bus
  • 1792 SPs, 112 TUs, 32 ROPs, 256-bit memory bus
  • 1920 SPs, 120 TUs, 32 ROPs, 256-bit memory bus

Of course, these specifications may be wrong as well…

Nvidia did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: Keep in mind that this news-story is highly speculative. We do not know exact sizes of Maxwell’s SMM, GPC and the un-core, therefore, take everything with a grain of salt. Nvidia has never revealed a die shot of the GM107 probably for a reason. Otherwise, it would be relatively easy to model the GM204.

The post Nvidia GeForce GTX 800: what to expect from its specs? first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-geforce-gtx-800-what-to-expect-from-its-performance-and-specs/feed/ 11
Sales of graphics cards dropped significantly in Q2 – report https://www.kitguru.net/components/graphic-cards/anton-shilov/sales-of-graphics-cards-dropped-significantly-in-q2-report/ https://www.kitguru.net/components/graphic-cards/anton-shilov/sales-of-graphics-cards-dropped-significantly-in-q2-report/#comments Tue, 17 Jun 2014 22:50:30 +0000 http://www.kitguru.net/?p=198676 Sales of graphics cards significantly dropped in the second quarter this year, according to a media report. In fact, shipments of discrete desktop graphics processing units have been declining for years now, but this time the drop of demand is attributed to crypto-currency miners. While the actual sales figures of discrete graphics cards for desktop …

The post Sales of graphics cards dropped significantly in Q2 – report first appeared on KitGuru.]]>
Sales of graphics cards significantly dropped in the second quarter this year, according to a media report. In fact, shipments of discrete desktop graphics processing units have been declining for years now, but this time the drop of demand is attributed to crypto-currency miners.

While the actual sales figures of discrete graphics cards for desktop PCs are yet to be revealed, DigiTimes web-site reports that shipments in the Q2 2014 were down 30 – 40 per cent compared to the first quarter of this year. According to Jon Peddie Research, total shipments of graphics cards in the Q1 2014 dropped to 14 million units, a decline of 0.8 per cent compared to the same quarter a year ago. In case the drop of demand is so significant, this may be the worst decline in GPU sales in the recent years.

The reason for the shipments drop suggested by sources in Taiwan is clearly not a trivial one. Apparently, miners of various crypto-currencies, such as Bitcoin, who primarily used AMD Radeon graphics cards to mine in the recent years, nowadays either move on to more energy-efficient ASICs designed specifically for mining, or simply quit the business and sell off their graphics cards via various Internet auctions. As a consequence, they do not buy new graphics boards and they affect the market by selling their old equipment on the Ebay.

amd_radeon_r9_290_1

Manufacturers of graphics adapters have reportedly asked AMD and Nvidia to cut-down the prices of the GeForce and the Radeon graphics processing units to make actual graphics cards more competitive, but both GPU vendors simply reduced their shipments.

At present it is unclear how many graphics cards are in the channel. If the are too lot of them, expect local retailers to cut-down their prices eventually. In case the demand and supply is balanced, price-cuts are unlikely.

AMD, Nvidia and graphics cards makers did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: The market of graphics cards is rather complex and cannot be affected by only one reason. Probably, there are a number of factors that influenced the market this way in the second quarter. In fact, traditionally, shipments of discrete graphics cards are down in the Q2 compared to the Q1.

The post Sales of graphics cards dropped significantly in Q2 – report first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/sales-of-graphics-cards-dropped-significantly-in-q2-report/feed/ 1
Nvidia already testing GM204-based graphics cards internally https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-already-testing-gm204-based-graphics-cards-internally/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-already-testing-gm204-based-graphics-cards-internally/#comments Mon, 05 May 2014 22:53:56 +0000 http://www.kitguru.net/?p=191273 Although not a lot of people outside of Nvidia Corp. know exactly what the code-named GM204 graphics processor unit is and when it is supposed to hit the market, it looks like there is one thing that we can be sure about: internal testing of the new graphics solution has already begun. A shipping declaration …

The post Nvidia already testing GM204-based graphics cards internally first appeared on KitGuru.]]>
Although not a lot of people outside of Nvidia Corp. know exactly what the code-named GM204 graphics processor unit is and when it is supposed to hit the market, it looks like there is one thing that we can be sure about: internal testing of the new graphics solution has already begun.

A shipping declaration of one of the packages sent by Nvidia to its test labs in India was partly published in Chinese forum at Chiphell web-site. The authenticity of the document could not be verified, so, it could be a fake, especially keeping in mind that it contains multiple typos. The package, which contains a test version of the GM204-based graphics card, was allegedly shipped in late April from China.

The test graphics card sent by Nvidia to its subsidiary in India is reportedly based on the code-named GM204 graphics processing unit and carries 8GB of GDDR5 memory. The declaration naturally lacks any specifications of the actual graphics processing unit, but 8GB of GDDR5 memory clearly points to an enthusiast-class graphics card. Since the board is based on alleged A01 revision of the chip, it is an early sample of the GM204-based product that will hardly ever reach the market.

nvidia_declaration_gm204

While we do know that the GM204 product is based on Maxwell architecture, we still do not known whether the chip is made using 20nm or 28nm process technology.

A rumour published at VideoCardz web-site indicates that the GM204-based graphics cards will be at least 10 per cent more powerful than the GeForce GTX 780 Ti, which means that the number of stream processors (SPs) inside the GM204 is similar or higher compared to GK110’s 2880 SPs. In case it is, it is likely that the chip is made using 20nm fabrication process.

Nvidia did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: In case the information about internal tests of the GM204 is correct and Nvidia is testing graphics cards based on the GM204 now, it is logical to expect commercial graphics cards powered by the chip to emerge on the market sometimes in late 2014. At least, the GM204 has good chances to become a 2014, not a 2015 product.

The post Nvidia already testing GM204-based graphics cards internally first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-already-testing-gm204-based-graphics-cards-internally/feed/ 2
Nvidia to tape-out mainstream, performance Maxwell GPUs this month – rumour https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-to-tape-out-mainstream-performance-maxwell-gpus-this-month-rumour/ https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-to-tape-out-mainstream-performance-maxwell-gpus-this-month-rumour/#comments Tue, 22 Apr 2014 12:05:04 +0000 http://www.kitguru.net/?p=188472 The less credible data about Nvidia Corp.’s next-generation graphics processing units is available, the more market rumours emerge. The recent set of unofficial pieces of information claims that Nvidia will tape-out its new-generation code-named GM204 (GM104) and GM206 (GM106) chips already this month, will manufacture them using 28nm process technology and will release them only …

The post Nvidia to tape-out mainstream, performance Maxwell GPUs this month – rumour first appeared on KitGuru.]]>
The less credible data about Nvidia Corp.’s next-generation graphics processing units is available, the more market rumours emerge. The recent set of unofficial pieces of information claims that Nvidia will tape-out its new-generation code-named GM204 (GM104) and GM206 (GM106) chips already this month, will manufacture them using 28nm process technology and will release them only in late-2014 – early-2015.

Due to low yields of chips made using 20nm fabrication technology at Taiwan Semiconductor Manufacturing Co. and some other reasons Nvidia has reportedly significantly reconsidered its plans regarding graphics processing units (GPUs) that were supposed to be made using 20nm manufacturing process, reports 3DCenter web-site. Instead of making its new GPUs based on the Maxwell architecture using 20nm technology, the company reportedly decided to indefinitely postpone them and roll-out yet another lineup of 28nm chips, it is reported.

As a result, the new-generation code-named GM204 (GM104) and GM206 (GM106) graphics processors (for performance and mainstream graphics cards, respectively) will be manufactured using current-gen 28nm fabrication process. The chips are reportedly set to be taped out (the point at which the artwork for the photomask of a circuit is sent for manufacturer) this month and emerge between December ’14 and February ’15.
It is unclear why the web-site that cites sources with knowledge of the matter calls first-generation Maxwell GPUs as GM206 and GM204, which are believed to be code-names of second-gen Maxwell graphics processing units. Previously both were known as GM100-series.

nvidia_artwork_iron

While the history knows examples when GPU vendors used current-gen/previous-gen process technology to produce graphics chips based on all-new architecture, there are not a lot of such examples. Usually new architectures mean new manufacturing processes as well. Moreover, in case chips are to be commercially produced in late 2014, it is highly likely that 20nm yields at TSMC will improve by that time.
It should be noted that in theory it is possible that Nvidia could make GM106 (GM206) and even GM104 (GM204) chips using 28nm fabrication process to completely replace all the GK106-based solutions as well as graphics boards powered by cut-down versions of the GK104.

To maintain healthy costs of the new GPUs, Nvidia would need to configure the new chips in such a way that they would be up to 360mm2 (GM104) and 240mm2 (GM106) large. Keeping in mind increased efficiency of Maxwell architecture compared to Kepler, such die sizes would ensure a noticeable performance boost compared to GK104- and GK106-based offerings, just what Nvidia needs for Holiday refresh cycle.

However, in case the GM204 (GM104) and GM206 (GM106) chips are to be made using 28nm process technology, Nvidia needed to make the design decision at least three quarters ago, or better earlier. Three quarters ago the company hardly had any idea regarding the state of 20nm node at TSMC in mid-2014. Besides, it probably was assured that its arch-rival AMD was hard at work on a 20nm product line and thus 28nm for nearly the whole family was not an option for Nvidia.

nvidia_artwork

It should be noted that the rumour about the GM104/GM204 chip made using 28nm fabrication process contradicts an earlier rumour which suggests that the new performance GPU from Nvidia will be more powerful than the code-named GK110 processor and will be made on 20nm node. It is impossible to make the GM104/GM204 faster than the GK110, manufacture it using 28nm process technology and sell it at lower price.
Nvidia did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: All the current rumours around the future GPUs should be taken with a grain of salt. What is interesting is that the talks about the performance GPU from the Maxwell lineup (whether it is called the GM104 or the GM204) are intensifying these days. Perhaps, the GPU is not that far away?

The post Nvidia to tape-out mainstream, performance Maxwell GPUs this month – rumour first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/nvidia-to-tape-out-mainstream-performance-maxwell-gpus-this-month-rumour/feed/ 1
First rumoured specs of GeForce GTX 880/GM104 published https://www.kitguru.net/components/graphic-cards/anton-shilov/first-rumoured-specs-of-geforce-gtx-880gm104-published/ https://www.kitguru.net/components/graphic-cards/anton-shilov/first-rumoured-specs-of-geforce-gtx-880gm104-published/#comments Fri, 11 Apr 2014 22:48:25 +0000 http://www.kitguru.net/?p=187089 Many sources from the GPU industry indicated last month that neither Advanced Micro Devices, nor Nvidia Corp. plan to make major GPU introductions at least till late in 2014, or maybe even early 2015. However, a Czech web-site this week not only revealed what it claims to be the specs of the GeForce GTX 880 …

The post First rumoured specs of GeForce GTX 880/GM104 published first appeared on KitGuru.]]>
Many sources from the GPU industry indicated last month that neither Advanced Micro Devices, nor Nvidia Corp. plan to make major GPU introductions at least till late in 2014, or maybe even early 2015. However, a Czech web-site this week not only revealed what it claims to be the specs of the GeForce GTX 880 graphics card and the GM104 chip, but also reported that Nvidia plans to introduce the new solution this summer.

Possible specs of the GM104 revealed

The GM104 graphics processing unit, which will succeed the almost legendary GK104 chip, is projected to feature whopping 3200 stream processors, 200 texture units and 40 raster operating units (a strange number!) as well as 256-bit memory controller. The GPU will consist of 7.9 billion of transistors and will be made using 20nm fabrication process, according to PCTuning web-site.

It is reported that graphics cards based on the GM104 will feature dual-slot cooling systems and will carry 2GB or 4GB of GDDR5 memory clocked at 7.40GHz. It is also expected that the graphics chip on the reference boards will operate at 900MHz clock-rate.

nvidia_geforce_gtx680_kepler

Nvidia GeForce GTX 680

It is interesting to note that the reference graphics card based on the GM104 is projected to sport only one 6-pin PCI Express power connector, which means that its maximum power consumption will be 150W (75W over PCIe slot and 75W over power cable).

Specifications of the GM104 GPU could not be confirmed with independent sources at press time, so we decided to analyse some of Nvidia’s past GPUs to figure out what the GM104 should be like.

Die sizes and transistor counts

Based on the specifications of different Nvidia chips released in the last five years, die sizes of the chips positioned for the same market segments remained similar:

  • High-end: 529mm2 – 571mm2 (the GT200b is not a good example of a performance chip since this is a shrink of another GPU that was developed just to cut-down costs);
  • Performance: 294mm2 – 332mm2 (the G92b is not a good example of a performance chip since it is also a die shrink);
  • Mainstream: 221mm2 – 238mm2;
  • Low-end: 116mm2 – 118mm2;

nvidia_gpu_specs

Architectures, process technologies and other things change over time, but die sizes of the chips remain on the same level. While there are exceptions, they just prove the overall “rule”. With that said, we can guess die sizes of the upcoming GPUs that belong to the Maxwell family.

Transistor count has consistently increased in accordance with the Moore’s law. In general, every new generation more than doubled the amount of transistors for the high-end chips, but that was not always true for performance, mainstream and low-end GPUs due to various reasons. While there is some kind of dependence of the transistor count of different GPUs within one family on the transistor count of the flagship chip, it is not really consistent across different GPU lines, so guesstimates regarding the Maxwell here can be absolutely wrong.

nvidia_maxwell_estimated_specs

Still, what we do know is that a chip featuring 7.9 billion of transistors should be made using 20nm process technology in order to hit the same price-points as the GK104 did. It is also pretty logical to expect the flagship GM100/GM110 chip to be made using 20nm fabrication process. At the same time, it is possible that Nvidia could make GM106 chip using 28nm fabrication process to completely replace all the GK106-based solutions as well as graphics boards powered by cut-down versions of the GK104.

20nm: ready or not?

There are a lot of signs that Nvidia needs TSMC’s 20nm fabrication process to complete the Maxwell lineup and show a huge performance bump. Of course, there is also a possibility that Nvidia could just develop Maxwell-based GPUs while maintaining existing transistor counts and die sizes. This could give Nvidia GPUs a speed bump while retaining their current costs. However, this is generally an unlikely scenario.

Only recently it was reported that the yield of 20nm chips at TSMC was around 50%, which is too low, keeping in mind that nowadays the world’s largest maker of semiconductors charges its customers on per-wafer, not per chip basis. Hence, until the yield issues are solved, it does not make a lot of sense for Nvidia to make chips using the 20nm manufacturing technology.

Summing up

All-in-all, while it is possible that the GM104 (and GeForce GTX 880) will indeed be made using 20nm fabrication process, consist of 7.9 billion of transistors and contain 3200 stream processors (thus beating the GeForce GTX Titan Black in many applications), chances that it will be released this summer are pretty thin, unless there is a major development with yields at Taiwan Semiconductor Manufacturing Co.

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

KitGuru Says: As it usually happens with the rumours, they can miss important pieces and the reality may turn out to be completely different than the logic tells us now.

The post First rumoured specs of GeForce GTX 880/GM104 published first appeared on KitGuru.]]>
https://www.kitguru.net/components/graphic-cards/anton-shilov/first-rumoured-specs-of-geforce-gtx-880gm104-published/feed/ 1