1. Introduction2. ADATA XPG V2 3100MHz Memory3. Configuring the BIOS4. System Configuration5. SiSoft Sandra6. MaxxMem27. Cinebench R11.5 64 bit8. Super Pi9. Performance Vs Avexir Core Extreme 3000MHz10. Overclocking11. Closing Thoughts12. View All Pages

Targeting extreme enthusiasts with recording-breaking overclock aspirations, ADATA's 3100MHz XPG V2 memory kit is designed to be the best of the best when it comes to raw frequency. Can the 3.1GHz kit's performance and overclocking potential prove that it currently sits at the top of the enthusiast memory pecking order?

With an out-of-the-box DRAM frequency of 3.1GHz, ADATA's XPG V2 3100MHz memory kit is currently the market's joint-fastest (Avexir also has a 3100MHz kit available, and a 3200MHz set which is scarce and of almost non-existent availability). If it's record-breaking overclock results you're after, then ADATA's 3100MHz memory kit gives a solid foundation for high-speed memory operations. A 2oz copper, 8-layer PCB and direct-touch heatsink add to the 3100MHz memory kit's features.

ADATA told us that the screening process to find the chips capable of hitting 3100MHz CL12 at 1.65V is ‘really hard'. Will that demanding screening process translate into some of the highest quality, and therefore most overclockable, chips currently available?

Specifications:

• PC3-24800 (3100MHz).
• Timings: 12-14-14-36.
• Voltage: 1.65V.
• Dual Channel Kit: (2x4GB).
• XMP 1.3 Ready.
• Lifetime Warranty.
• Model number: AX3U3100W4G12-DGV.

The ADATA XPG V2 3100MHz memory sticks are shipped in a basic clamshell package. Kit capacity and speed is outlined on the packaging's front side, but not the timings – you'll need to read the specifications sticker for those.

Unlike G.Skill's 2933MHz TridentX memory, ADATA's XPG V2 3100MHz kit does not ship with a fan. Given that the ADATA's kit's target audience is likely to be using extreme cooling, a RAM fan isn't entirely necessary, but it would have been convenient to users testing air overclocking performance.

A more detailed specifications sticker is placed on the packaging's rear side. The plastic clamshell also utilises an ‘easy-open' design which relies on simply pulling a section of the top-left corner, not having to going crazy with a pair of scissors.

Note: As the sticks are identical in all respects, except for their operating frequencies, the following written analysis has been taken from our review of the 2800MHz version of ADATA's XPG V2 kit and re-used.

The ADATA XPG V2 memory modules are covered by a uniquely-shaped, gold heatsink that is designed to transfer heat away from the memory chips, and add an attractive touch to the modules. With Asus’ range of Z87 motherboards utilising gold heatsinks, matching the XPG V2 memory with other components is a possibility.

Giving a capacity of four gigabytes per DIMM for our 8GB kit, eight memory chips are situated on one side of the PCB, with a spacer occupying the other.

Measuring in at a maximum of 43mm-tall, the gold heatsink of ADATA’s XPG V2 memory modules features a finned design that is intended to distribute heat away from the PCB and chips. The heatsink is made of two separate parts which are then united by virtue of their mounting locations and a small clip.

As well as the gold kit that we received, ADATA’s XPG V2 memory is also available with a tungsten grey-coloured heatsink.

ADATA puts its brand name and the XPG product logo on either side of the memory modules’ black PCBs. The XPG V2 memory kit on show today is one of the fastest sets in the world. It operates at a DRAM frequency of 3100MHz with timings of 12-14-14-36.

If required, each heatsink can be removed with relative ease, allowing the black PCB to exist in a less-interfering form factor.

We are using the ASRock Z87 OC Formula motherboard to test today's memory kit. It supports RAM from many manufacturers and gives us a large amount of flexibility to tinker with memory multiplier, base clock and timings settings. It has also reached the highest DRAM frequency of all the Z87 motherboards that we have tested.

Our full review of the ASRock Z87 OC Formula motherboard can be read here.

As Haswell supports a maximum memory speed of 2933MHz using the 1.00x CPU strap, ADATA opted to use the 1.25x CPU strap to reach 3100MHz. The base clock is set as 129.2MHz which, when used with the 100:133 BCLK:DRAM ratio, provides 3100MHz by using the 18x memory multiplier.

The fully-functioning XMP profile configured the memory kit for use at its 3100MHz frequency and correct timings.

Windows and CPU-Z registered the memory's functioning settings as correct.

Our system's CPU-Z validation running at a 3100MHz DRAM frequency can be found here. Notice the sub-standard CPU frequency? Our system opts for a CPU multiplier that kept our 4770K operating below its standard stock speed, let alone the frequency that most motherboards' forced turbo profiles configure.

In light of our ASRock- and ADATA-based system's (wise, in all honesty) decision to opt for a low multiplier that provided a CPU frequency of 3.36GHz when XMP was enabled, we forced a higher core ratio so that our processor would sit as close to 3900MHz as possible. This will help us minimise the variation in performance results against competing kits which used a CPU frequency of 3.9GHz (forced turbo on the 4770K).

The validation of the memory frequency (3100MHz) and CPU speed (3875MHz) that we used for testing can be viewed here.

We are using an Intel Core i7 4770K processor with an ASRock Z87 OC Formula motherboard.

The ADATA XPG V2 memory kit's stock speed is: 3100MHz 12-14-14-36-2T @ 1.65V.

Note: CPU and CPU Cache frequencies for the Avexir Core Extreme Series 3000MHz and ADATA XPG V2 3100MHz memory sets are different to the speeds set for the comparisons kits. This is due to the Avexir and ADATA memory’s use of out-of-spec BCLK frequencies which then result in non-standard CPU and CPU Cache speeds. Due to this fact, the results and charts should be viewed with care, and our performance analysis should be read to make the results clearer.

Z87 Test System:

• Processor: Intel Core i7 4770K (set to 3875MHz via a 129.15MHz BCLK and 30x core ratio).
• Motherboard: ASRock Z87 OC Formula
• Graphics Card: nVidia GTX 760 2GB.
• System Drive: 500GB Samsung 840 Series SSD.
• CPU Cooler: Corsair H100i.
• Case: NZXT Phantom 630.
• Power Supply: Seasonic Platinum 1000W.
• Operating System: Windows 7 Professional 64-bit.

Comparison memory:

• 8GB (2x 4GB) Patriot Viper Xtreme Division 2 2133MHz (11-11-11-30).
• 8GB (2x 4GB) Avexir Core Extreme Series 3000MHz (12-14-14-35).
• 16GB (2x 8GB) Corsair Vengeance Pro 2400MHz (10-12-12-31).
• 8GB (2x 4GB) ADATA XPG V2 2800MHz (12-14-14-36).

Software:

• MaxxMem 2.
• SiSoft Sandra 2013.
• Super Pi.
• Cinebench R11.5 64 bit.

SiSoftware Sandra (the System ANalyser, Diagnostic and Reporting Assistant) is an information & diagnostic utility. It should provide most of the information (including undocumented) you need to know about your hardware, software and other devices whether hardware or software.

Sandra is a (girl’s) name of Greek origin that means “defender”, “helper of mankind”. We think that’s quite fitting.

It works along the lines of other Windows utilities, however it tries to go beyond them and show you more of what’s really going on. Giving the user the ability to draw comparisons at both a high and low-level. You can get information about the CPU, chipset, video adapter, ports, printers, sound card, memory, network, Windows internals, AGP, PCI, PCI-X, PCIe (PCI Express), database, USB, USB2, 1394/Firewire, etc.

Native ports for all major operating systems are available:

• Windows XP, 2003/R2, Vista, 7, 2008/R2 (x86)
• Windows XP, 2003/R2, Vista, 7, 2008/R2 (x64)
• Windows 2003/R2, 2008/R2* (IA64)
• Windows Mobile 5.x (ARM CE 5.01)
• Windows Mobile 6.x (ARM CE 5.02)

All major technologies are supported and taken advantage of:

• SMP – Multi-Processor
• MC – Multi-Core
• SMT/HT – Hyper-Threading
• MMX, SSE, SSE2, SSE3, SSSE3, SSE 4.1, SSE 4.2, AVX, FMA – Multi-Media instructions
• GPGPU, DirectX, OpenGL – Graphics
• NUMA – Non-Uniform Memory Access
• AMD64/EM64T/x64 – 64-bit extensions to x86
• IA64 – Intel* Itanium 64-bit

Due to its loose primary and secondary timings, ADATA's 3100MHz XPG V2 kit makes a slow start in Sandra's memory bandwidth test. We were also surprised to see ADATA's identical latency 2800MHz outperform the 3100MHz variant; the CPU speed deficit of 25MHz may be to account for the 3100MHz set's bottom-place finish.

The cache bandwidth result takes ADATA's 3100MHz set off the bottom of the chart and places its performance very close to that of Avexir's 3GHz kit. Bear in mind that Avexir's setup has a circa-100MHz CPU speed advantage over the other contenders – a point which may have a small effect on the results.

The high speed memory kits that we have tested continue to impress us with their latency performance, as pointed out by Sandra's test. The 3100MHz 12-14-14-36-2T XPG V2 kit from ADATA tops the table when it comes to cache and memory latency. 

MaxxMem2 PreView is a handy, free program to rate memory performance. It can be downloaded over here.

Throughput performance isn't strong for ADATA's 3100MHz memory with loose timings. Again, the 25MHz CPU speed deficit, in comparison to the 2800MHz set, forces ADATA's 3100MHz memory to sit firmly at the bottom of Maxxmem2's throughput chart.

The latency performance, as displayed by Maxxmem2, is very strong for the 3100MHz CL12 memory sticks. ADATA's flagship kit shows latency performance that is around 1.3% better than that of Avexir's 3000MHz set, despite the latter's higher CPU frequency. Going from a 2800MHz memory to 3100MHz, using identical timings, garners a latency reduction of around five-and-a-half percent.

CINEBENCH R11.5 64 Bit is a real-world cross platform test suite that evaluates your computer’s performance capabilities. CINEBENCH is based on MAXON’s award-winning animation software CINEMA 4D, which is used extensively by studios and production houses worldwide for 3D content creation. MAXON software has been used in blockbuster movies such as Spider-Man, Star Wars, The Chronicles of Narnia and many more.

CINEBENCH is the perfect tool to compare CPU and graphics performance across various systems and platforms (Windows and Mac OS X). And best of all – it’s completely free.

There's little point reading too much into the Cinebench score registered by ADATA's 3100MHz memory kit; Cinebench is heavily CPU-weighted and penalises the 3100MHz ADATA system for its 25MHz lower CPU clock speed.

Super Pi is used by a huge audience, particularly to check stability when overclocking processors. If a system is able to calculate PI to the 2 millionth place after the decimal without mistake, it is considered to be stable in regards to RAM and CPU.

We used Super Pi’s ’16M’ benchmark setting.

Despite being paired with a 25MHz lower CPU frequency, ADATA's 3100MHz memory sticks register a score that puts them as the test's joint leader. One second is shaved off the time set by ADATA's almost identical 2800MHz memory.

Due to their out-of-spec base clock configurations, the Avexir Core Extreme Series 3000MHz and ADATA XPG V2 3100MHz memory kits were set to operate at different CPU frequencies, when the default core ratios are used.

For the following tests, we compared both of the ultra-high speed – 3GHz+ – memory kits against each other at almost identical CPU frequencies.

We tweaked the CPU multiplier to get each kit operating with processor speeds as close as possible to each other. This should give an accurate comparison of the performance gains on offer (if any) for higher speed memory.

ADATA XPG V2 3100MHz configuration: 3100MHz 12-14-14-36-2T @ 1.65V (with 4003MHz CPU frequency).

Avexir Core Extreme Series 3000MHz configuration: 3000MHz 12-14-14-35-2T @ 1.65V (with 3988MHz CPU frequency).

Despite operating at a higher DRAM frequency, ADATA's 3100MHz kit still loses out to Avexir's 3000MHz set in Sandra's memory bandwidth test. Given that both kits operate at almost identical primary timings, the performance deficit displayed by the ADATA memory could be related to secondary timings, or even systematic and measurement errors in Sandra's benchmark.

With its slightly-faster CPU frequency and 100MHz DRAM speed increase, the ADATA's kit manages to offer 1.7% more cache bandwidth than Avexir's 3000MHz sticks.

Latency performance favours the faster kit, given that both set's primary timings are almost identical. ADATA's 3100MHz memory shows a consistent 2.2% performance lead in both latency tests.

Despite its greater DRAM speed, the Maxxmem2 benchmark does not like the configuration of ADATA's 3100MHz memory. It's hard to think of anything other than secondary timings that could be to blame for Avexir's lower-speed kit with near-identical timings offering greater performance.

Other than one-hundredth of a point in Cinebench (around 0.1% – well within error margins), there is nothing between the 3000MHz and 3100MHz memory kits when it comes to real-world performance.

With memory modules of this calibre, it is fair to assume that a large proportion of the user base will be looking for the kit's highest attainable frequency with good enough stability. That doesn't mean that the ability for the overclocked kit to run every benchmark with stability is crucial, but it does place an importance on being able to boot and run at least one test flawlessly.

We will be tweaking the base clock to fine-tune the memory overclock. Due to the above point, and the fact that the processor frequency will be different to when the RAM was running at stock speeds, there is little point in showing comparison benchmarks. Instead, we will use a single run of Super Pi 16M to verify stability.

As we do not have access to sub-zero testing equipment, we will be relying on tweaks and voltage settings when using an air-cooled system.

We started out by using the 1.25x CPU strap so that we could use a BCLK above the 125MHz-mark for our overclocking attempts. We soon realised that the ADATA XPG V2 modules had plenty of overclocking headroom and we therefore switched to using the 1.67x CPU strap.

To eliminate base clock bottlenecks, we used a number of increased voltages for the CPU and system. The CPU multiplier was dropped to eliminate instability due to a high processor speed.

The tweaks applied included a 1.7015V DRAM voltage, various CPU voltage increases and setting changes for BCLK overclocking stability, and a CPU input voltage of 1.900V. ASRock’s automated settings proved favourable for overclocking performance and didn’t require changing.

Our best result using the stock timings was 3280MHz (164MHz BCLK) using a DRAM voltage of 1.7015V. We did manage to reach the same 3288MHz (using the 1.25x CPU strap and a 137MHz BCLK) that Avexir's Core Extreme Series modules hit using the ASRock Z87 OC Formula motherboard, but stability in Windows was certainly questionable, as proven by two consecutive BSODs.

Our validation with the ADATA XPG V2 3100MHz memory kit running at a 3280MHz DRAM frequency with stock timings of 12-14-14-36-2T can be viewed here.

In order to push further and obtain the maximum memory frequency at 1.7015V, timings were loosened to 14-15-15-40-2T.

The highest memory overclock that we achieved with the ADATA XPG V2 memory modules was 3340MHz, using a 167MHz base clock.

We wouldn't call this a 24/7 stable overclock with our settings, although a little voltage tweaking (with better cooling, perhaps) would have garnered improved stability.

3340MHz was the highest frequency that we achieved; 1MHz higher on the BCLK and the system simply would not boot.

Our validation running at a 3340MHz DRAM frequency with loosened timings can be viewed here.

ADATA's 3100MHz XPG V2 modules operate at a blisteringly-fast 3.1GHz and have the overclocking headroom to take them past 3.3GHz on air, with a sensible voltage. The hand-picked, high-quality memory chips used on ADATA's XPG V2 modules have given a clear indication that they have the potential to reach ultra-high overclocks with extreme cooling and higher voltage levels than the 1.7015V we used.

While the stock performance may not be the best in every circumstance – a fact that is largely due to the loose timings – nobody buying this set of memory is going to leave the sticks operating in their default states. Adjusting the primary and secondary timings will garner performance improvements that allow the 3.1GHz DRAM frequency to show its performance potential.

Despite performance improvements being easy to obtain through tweaking, leading benchmark results aren't this kit's primary goal. As proven by our test comparisons, many lower-speed and tighter-latency memory kits are able to outperform modern ultra-high frequency kits. Record-breaking frequency results are what ADATA's 3100MHz XPG V2 memory kit is designed for, and as far as our overclocking results go, they're en-route to those achievements.

By simply increasing the DRAM voltage from 1.65V to 1.7015V and tweaking some motherboard settings, we managed to boost the ADATA XPG V2 3100MHz memory kit's DRAM frequency to 3280MHz. Increasing the speed of an already-fast kit by almost 6% with very little effort is a positive point that gives an indication of the memory chips' potential. With relaxed timings we hit 3340MHz – a frequency boost of just under 8%.

Motherboard compatibility is unlikely to be a concern with the ADATA XPG V2 3100MHz memory kit. The XMP profile makes use of a 129.2MHz BCLK which adjusts the CPU strap to 1.25x. Many of the enthusiast Haswell processors should be able to hit 3100MHz without problems, especially when using the 1.25x or 1.67x CPU strap. Motherboard limitations are unlikely to cause issues; we are certain that the type of user buying this memory kit will already have one of the best overclocking-geared motherboards available.

ADATA ships the XPG V2 memory kit with gold or tungsten-grey heatsinks. While either colour may not appeal to all, it is highly likely that this kit's default heatspreaders will be removed in favour of extreme heatsinks that allow for a greater cooling capacity.

When we asked ADATA for the 3100MHz kit's cost shortly before processing this review, we were told that the current pricing is unstable (presumably the repercussions of recent fires at Hynix plants). A quick look on Newegg sees the 3100MHz XPG V2 8GB memory kits selling for a card-crunching $999.99. If this pricing holds firm, we can expect the kit to retail for around £700-800 in the UK.

While almost a thousand dollars is an obscene amount for the wide majority of people to even consider spending on eight gigabytes of RAM, ADATA has a very specific target audience for its 3100MHz memory modules. That customer group is one that has no limits on what hardware costs, provided it is the best available. To a record-breaking overclocker or extreme enthusiast, spending almost $1000 to get the world's fastest memory kit is a possibility.

When it comes to breaking world records for frequency numbers, for the professionals, nothing but the best will do. With that in mind, ADATA's 3100MHz XPG V2 can call itself one of the world's fastest memory kits and will grab much attention from extreme users looking to break frequency records.

Pros:

• One of the fastest memory kits in the world, out-of-the-box.
• 3100MHz XMP profile.
• Good overclocking headroom – can hit a loose 3340MHz on air with 1.7015V.
• Memory chips hand-picked for their high frequency performance.

Cons:

• Very, very expensive.
• A fan would be good for air overclocking scenarios.
• Best for ultimate frequency, not necessarily ultimate performance.

KitGuru says: One of the fastest sets of memory in the world, ADATA's tirelessly-screened XPG V2 3100MHz kit is worth buying if you're an extreme overclocker with aspirations to push frequency boundaries and break records.

Become a Patron!