Deus Ex: Mankind Divided is the latest AAA title to support DirectX 12, with its developer Eidos deploying a DirectX 12 renderer weeks after its release, through a patch. Guru3D put the DirectX 12 version of the game through five GPU architectures, AMD "Polaris," GCN 1.1, GCN 1.2, NVIDIA "Pascal," and NVIDIA "Maxwell," through Radeon RX 480, Radeon R9 Fury X, Radeon R9 390X, GeForce GTX 1080, GeForce GTX 1060, and GeForce GTX 980. The AMD GPUs were driven by RSCE 16.9.1 drivers, and NVIDIA by GeForce 372.70.

Looking at the graphs, switching from DirectX 11 to DirectX 12 mode, AMD GPUs not only don't lose frame-rates, but in some cases, even gain frame-rates. NVIDIA GPUs, on the other hand, significantly lose frame-rates. AMD GPUs tend to hold on to their frame-rates at 4K Ultra HD, marginally gain frame-rates at 2560 x 1440, and further gain frame-rates at 1080p. NVIDIA GPUs either barely hold on to their frame-rates, or significantly lose them. AMD has on multiple occasions claimed that its Graphics CoreNext architecture, combined with its purist approach to asynchronous compute make Radeon GPUs a better choice for DirectX 12 and Vulkan. Find more fascinating findings by Guru3D here.More graphs follow.

AMD today released its third Radeon Software Crimson Edition update for this month. Version 16.5.2.1 provides a massive 35 percent performance increase for Radeon R9 390 series graphics cards (R9 390, R9 390X) on "Doom" (2016), compared to version 16.5.2, released earlier this month. The change-log does not mention whether this is also applicable to the nearly identical Radeon R9 290 series (R9 290, R9 290X, and R9 295X2). Not much else has changed, which explains the version numbering.DOWNLOAD: AMD Radeon Software Crimson Edition 16.5.2.1 for Windows 10/8.1/7 64-bit | Windows 10/8.1/7 32-bit

AMD, which has been timing its upcoming dual-GPU "Fiji" graphics card to launch sometime this year, may have demoed a production version of the card in one of its launch partners, Falcon Northwest's, Tiki high-end gaming desktop, as a "VR developer box." AMD's Roy Taylor, in a recent tweet, captions a picture of this dev box as being "the world's best DirectX 12 VR developer box," leading the press to speculate that it's running the company's dual-GPU "Fiji" card.

A close look at AMD's VR dev box, through its windowed graphics card compartment, reveals an air-cooled AMD reference graphics card, which VideoCardz' trigonometry pins as being shorter than a Radeon R9 390X reference board. It could be a reference R9 380X, but then a reference dual-GPU "Fiji" PCB is roughly of the same length, and a R9 380X wouldn't earn the title of being the "world's best" from a senior AMD exec while there are faster AMD cards, such as the R9 Fury. The ability of the full-spec "Fiji" silicon to cope well with a rather simple air-cooler in the R9 Nano fans even more speculation that a dual-GPU "Fiji" board could make do with a powerful air-channel cooler.

AMD gave its premium small-factor gaming graphics card, the Radeon R9 Nano, its first major price cut. The card now starts at US $499, down from its launch price of $649. At $499, the R9 Nano is priced on par with its similar-performing albeit bigger and noisier sibling based on the "Fiji" silicon, the Radeon R9 Fury. The company's flagship single-GPU card, the R9 Fury X, remains at $599, its price was gradually reduced from its launch price of $649.

The three SKUs appear to be positioned to compete with NVIDIA's GeForce GTX 980, and offer cost-effective alternatives to the $629 GTX 980 Ti. Elsewhere in the lineup, the Radeon R9 390X starts at $379, and has its guns trained on the GTX 980 and GTX 970. Its smaller sibling, the Radeon R9 390 starts at $299.

Stardock's "Ashes of the Singularity" may not be particularly pathbreaking as an RTS, in the Starcraft era, but has the distinction of being the first game to the market with a DirectX 12 renderer, in addition to its default DirectX 11 one. This gave gamers the first peak at API to API comparisons, to test the tall bare-metal optimizations of DirectX 12, and as it turns out, AMD GPUs do seem to benefit big.

In a GeForce GTX 980 vs. Radeon R9 390X comparison by PC Perspective, the game seems to perform rather poorly on its default DirectX 11 renderer for the R9 390X, which when switched to DirectX 12, not only takes a big leap (in excess of 30%) in frame-rates, but also outperforms the GTX 980. A skeptical way of looking at these results would be that the R9 390X isn't optimized for the D3D 11 renderer to begin with, and merely returns to its expected performance vs. the GTX 980, with the D3D 12 renderer.

TUL Corporation, a leading and innovative manufacturer of AMD graphics cards since 1997, has released the Devil R9 390X 8GB GDDR5. It is based on the latest GCN architecture to help deliver outstanding and extraordinary graphics performance and image quality.

Devil R9 390X utilizes 8GB of GDDR5 memory, 2816 stream processors, comes with a core clock speed at 1100MHz, and 1525MHz memory clock speed which is connected via a new high speed 512-bit memory interface. For enhancing power efficiency and preventing losses, Digital PWM solution is provided to work at a higher frequency in order to support fine tuning adjustment and low ripple at an output voltage. Moreover, the total of 8 phases (6+1+1) board design is applied to the product for power efficiency, stability, and delivering ultimate performance at OC mode. PowIRstage increases the power up to 3-13%, features up to 1.0 MHz switching frequency, and supports efficient cooling ability. This model supports Direct 12 and AMD's newest technologies such as Virtual Super Resolution, FreeSync, Liquid VR, and 4K resolution.

GIGABYTE introduced a pair of custom-design Radeon R9 390 series graphics cards based on its new 2-slot WindForce 2X cooling solution, the R939WF2-8GD (R9 390) and the R939XWF2-8GD (R9 390X). The two are essentially the company's R9 390 series G1.Gaming products without the factory-overclock. Both cards are identical to their G1.Gaming siblings in every other way, except the branding and box-art. The two cards stick to AMD reference clock speeds, and offer 0 dBA idle cooling. The two should be cheaper than the G1.Gaming products.

VisionTek Products LLC, a leading manufacturer of award-winning, high-performance upgrades and accessories for PCs and Macs, today announced the Radeon Fury X, alongside five other new Radeon R9 300 and R7 300 graphics cards designed with Advance Micro Devices (AMD) GPUs. The new VisionTek Radeon graphics cards will feature a new extreme look and are available starting June 24th, 2015.

The VisionTek Radeon Fury X opens a whole new dimension of gaming, for an unbelievable alternate reality. It is the world's first graphics card with AMD-pioneered High Bandwidth Memory (HBM) integrated on-chip, delivering 60% more memory bandwidth over GDDR5. The VisionTek Radeon R9 Fury X features black-nickel aluminum exoskeleton, maintenance-free AIO closed-loop liquid cooling, GPU Tach activity indicator, and red LED "RADEON" illumination along the spine of this wicked beast. The compact 7.5-inch card packs a performance punch, leading the way to create small and powerful PCs by condensing the HBM memory into 94% less space than GDDR5. The VisionTek Fury X was built to enable the next generation of 4K and VR gaming. With this card, PC users won't just upgrade, they will revolutionize.

ASUS managed to tame AMD's feisty "Grenada" silicon, which powers the Radeon R9 390 and Radeon R9 390X, by announcing two high-end graphics cards based on its new triple-fan STRIX DirectCU 3 cooling solution. The cooler turns its fans off when the GPU is idling (common desktop / light-3D loads), and begins to spool up only under heavy 3D loads. The company claims that this will be the quietest R9 390 series cards you can buy.

The STRIX DirectCU 3 cooler is the same as the one pictured cooling the GeForce GTX 980 Ti STRIX, which we spotted at Computex. It features a huge monolithic aluminium fin-stack heatsink, to which heat drawn from the GPU is fed by four 10 mm thick nickel-plated copper heat pipes. This heatsink is ventilated by three 100 mm spinners. This heatsink has contact bases even over the card's 8-phase VRM, and a base-plate that draws heat from its 16 GDDR5 memory chips, that make up 8 GB. The R9 390 STRIX offers factory OC of 1050 MHz (vs. 1000 MHz reference); while the R9 390X STRIX offers 1070 MHz (vs. 1050 MHz reference). The memory ticks at 6.00 GHz on both cards. ASUS didn't announce pricing.

GIGABYTE kicked off its Radeon R7 300 and Radeon R9 300 graphics card lines, by leveraging its latest generation of in-house design WindForce cooling solutions, and custom PCBs. All the cards launched by GIGABYTE today, are factory-overclocked; some even bear the company's flagship G1.Gaming branding. The series begins with the GV-R736OC-2GD, based on the R7 360, with factory overclocked speeds of 1200 MHz core (vs. 1050 MHz reference); and a single 90 mm fan based heatsink cooling solution. Next up, is the GV-R737WF2OC-2GD/4GD, which comes in 2 GB and 4 GB variants, factory OC of 1015 MHz (vs. 975 MHz reference), and GIGABYTE's compact rendition of its WindForce 2X cooling solution.

Next up, is the Radeon R9 380 G1.Gaming SOC (GV-R938G1 GAMING-4GD), featuring 4 GB of memory, a minor OC of 990 MHz (vs. 970 MHz reference), and a surprisingly compact mid-tier WindForce 2X cooler in charge of taming this 190W chip. We then move on to the big boys, the R9 390 G1.Gaming and R9 390X G1.Gaming, with which GIGABYTE again surprises us with 2-slot, 2-fan WindForce 2X cooling solutions, handling the 275W cards. The R9 390 G1.Gaming ticks at 1025 MHz (vs. 1000 MHz reference); while the R9 390X G1.Gaming runs its core at 1060 MHz (vs. 1050 MHz reference).

TUL Corporation, a leading and innovative manufacturer of AMD graphic cards since 1997, has released a whole new line of graphics cards that you have all been waiting for. All models are based on the latest GCN architecture to help deliver outstanding and extraordinary graphics performance and image quality. PCS+ R9 390X/390 both utilizes 8GB of GDDR5 memory with 2816 stream processors and ships with core clock speed at 1060MHz for the 390X, 1010MHz on the 390, and 1500MHz memory clock speed which is connected via a new high speed 512-bit memory interface. For enhancing power efficiency and preventing losses, Digital PWM solution is provided to work at a higher frequency in order to support fine tuning adjustment and low ripple at an output voltage. Moreover, the total of 8 phase (6+1+1) board design is applied to the models for power efficiency, stability, and delivering ultimate performance at OC mode.

PowlRstage increases the power up to 3-13%, features up to 1.0 MHz switching frequency, and supports efficient cooling ability. Also, the anodized back plate is attached to the back of the card to protect the components as well as helping to lower the temperature. The Efficient Trio Fan design with the Mute Fan Technology delivers intelligent fan controlling to provide noiseless environment as well as reducing the power consumption when the GPU temperature is lower than 60°C. For the best thermal solution, the card is equipped with pure copper GPU block to absorb heats from GPU. This block is connected to 3pcs of 6mm heat pipes and 1pc of 8mm heat pipe to help deliver heat to the nickel plated aluminum fins heat sink. These two models support AMD's newest technologies such as Virtual Super Resolution, FreeSync, Liquid VR, and 4K resolution.

Building on the success of the previous generation, the new series features the latest Graphics Core Next architectures from AMD, paired with larger frame buffers and the evolution of our award winning cooler solutions to deliver the very best performance ratings and support for the latest technologies including DirectX 12, AMD FreeSync, TrueAudio, Liquid VR, VSR scaling and support for Ultra HD. This new generation includes models with our industry acclaimed Tri-X and Dual-X cooling technologies and will be available from 18th June 2015.

At the top of the range is the new SAPPHIRE Tri-X R9 390X which features the latest version of our award winning Tri-X cooler. This now features dual ball bearings in each of the three fans for higher reliability and enhanced Intelligent Fan Control (IFC-II) which turns off the fans completely for silent operation under light load. Its heatpipe array with its industry leading 10mm heatpipe and diecast heatsink design keeps the card running cool even under the most demanding applications, and the use of long life capacitors and SAPPHIRE Black Diamond chokes contribute to consistent performance and high reliability.

Today, Club3D is proud to announce the brand new Club 3D Radeon R7 and R9 300 series! Value for money has always been of key importance for AMD based graphics cards and the new stack pushes the boundaries again, providing unprecedented FPS for your money! Whether you are looking for a capable new graphics card to dominate online gaming in Full HD 1080p resolution or to have the best performance per dollar the industry has to offer to play the latest AAA titles on your Ultra HD 2160p screen or multi monitor setup, rest assured that you will find the card that meets your demands in the new Radeon R7 and R9 300 series.

AMD wrapped up today's GPU launch marathon, with the Radeon R9 390 series; which includes the R9 390, and the R9 390X. The Radeon R9 390 is priced at US $329, and offers performance competitive to the GeForce GTX 970. The R9 390X, on the other hand, is starts at US $429, and offers performance that's between the GTX 970 and GTX 980, while being closer to the latter. Both are based on the 28 nm "Grenada" silicon, which is the "Hawaii" silicon re-hashed.

The R9 390 packs 2,560 stream processors, 160 TMUs, 64 ROPs; while the R9 390X offers 2,816 stream processors, 176 TMUs, and 64 ROPs. Both cards offer 8 GB of GDDR5 memory, across the chips' 512-bit wide memory interfaces. Both cards let you game at 1440p with settings maxed out; or 4K Ultra HD, with reasonably high eye-candy. The R9 390 features core clock speeds of 1000 MHz, while the R9 390X tops that with 1050 MHz core. The memory on both cards, is clocked at 6.00 GHz (GDDR5-effective), translating into a staggering 384 GB/s memory bandwidth.

In all the buzz surrounding the five products based on its Fiji silicon, AMD also announced five other mid-thru-performance segment graphics cards, the Radeon R7 360, the Radeon R7 370, the Radeon R9 380, the Radeon R9 390, and Radeon R9 390X. Aimed at competitive online MOBA gaming the Radeon R7 360 is good enough to play MOBAs such as "League of Legends," at 1080p, and most other modern games at 900p and 720p.

Based on the "Bonaire" silicon, the Radeon R7 360 features 768 stream processors, 48 TMUs, 16 ROPs, and a 128-bit wide GDDR5 memory interface, holding 2 GB of memory. The core is clocked at 1050 MHz, and the memory at 6.50 GHz (GDDR5-effective), translating into 104 GB/s of memory bandwidth. The card draws power from a single 6-pin PCIe power connector, and has a typical board power rating of 100W.

The Radeon R7 370 is designed for MOBA, FPS, and MMORPGs at 1080p resolution. It is expected to feature 1,024 stream processors, 64 TMUs, 32 ROPs, and a 256-bit wide GDDR5 memory interface, holding 2 GB or 4 GB of memory. The core is clocked at 975 MHz, and the memory at 5.40 GHz (GDDR5-effective), belting out 179 GB/s of memory bandwidth. AMD has given this chip some energy optimizations, which lends it a typical board power of just 110W. The card draws power from a single 6-pin power connector.

AMD's upcoming Radeon R9 390X and R9 390 performance-segment graphics cards reportedly feature higher GPU and memory clocks over the products they are a re-branding of, the R9 290X and R9 290, respectively. The 28 nm "Grenada" silicon they are based on, is identical to "Hawaii," down to the last transistor. This has been confirmed by leaked GPU-Z screenshots, which reveal the device-IDs of the two cards to be identical to those of the R9 290X and R9 290. Since the Device-IDs are the same, GPU-Z is reading the chip as "Hawaii." The code-name "Grenada" appears in the BIOS version string.

Unlike older, more blatant re-brands, such as GeForce 8800 GT to 9800 GT, AMD did drop in a few changes. To begin with, the memory amount has been doubled on both cards, to 8 GB. The memory clock has been increased from 1250 MHz (5.00 GHz GDDR5-effective), to 1500 MHz (6.00 GDDR5-effective), resulting in memory bandwidth increase to 384 GB/s, up from 320 GB/s. The core clock speed on the R9 390X is 1050 MHz (up from 1000 MHz on R9 290X); and 1000 MHz on the R9 390 (up from 947 MHz on the R9 290).

People with access to an XFX Radeon R9 390X graphics card, took it apart to take a peek at its PCB. What they uncovered comes as no surprise - the underlying PCB is identical in design to AMD reference PCB for the Radeon R9 290X, down the location of every tiny SMT component. At best, the brands on the chokes and bigger conductive polymer caps differ; and 512 Gbit GDDR5 chips under the heatspreader, making up 8 GB of the standard memory amount. The GPU itself, codenamed "Grenada," looks identical to the "Hawaii" silicon which drove the R9 290 series. It's highly unlikely that it features updated Graphics CoreNext 1.2 stream processors, as older rumors suggested.

Ahead of its possible June 16 launch, more pictures of AIB-branded Radeon R9 390X graphics cards are hitting the wires. Here, we have an XFX-branded R9 390X, complete with its box-art. The R9 390X, is expected to be a re-brand of the previous generation R9 290X, with its standard memory amount raised to 8 GB. It's based on the 28 nm "Grenada" silicon. We've seen no evidence pointing at "Grenada" being some sort of an upgrade of "Hawaii" with newer GCN 1.2 stream processors. Perhaps AMD polished its electricals to the extent it could, without changing the silicon. We'll know for sure only next week.

XFX' Radeon R9 390X features a custom air cooling solution, which is taller than the one the company used on its R9 290 series products. It still retains its 2-slot form. The cooler consists of two aluminium fin-stacks, along the edges of seven 8 mm thick copper heat pipes, which draw heat from the GPU at the base. A metal heatspreader conveys heat from the memory chips to the main heatsink; while individual metal heatsinks cool the VRM. The card draws power from a combination of 6-pin and 8-pin PCIe power connectors. Display outputs include two dual-link DVI, and one each of HDMI 1.4a and DisplayPort 1.2a connectors.

Johan Andersson, technical director for the Frostbite game engine at DICE, developers of games such as Battlefield, tweeted the first clear picture of AMD's next flagship graphics card, and it looks a lot [better] than this mockup render. We'd be tempted to call it the Radeon R9 390X, but older reports suggest that AMD could give it a fancy name, just as NVIDIA named its top-dog "Titan." That's not all, Andersson commented that "this new island is one seriously impressive and sweet GPU," referring to the card's GPU codename of "Fiji." AMD is expected to launch this card in the third week of June. Either to preempt that, or out of spook (with an effort to siphon off high-end GPU sales), NVIDIA is preparing the GeForce GTX 980 Ti, which will launch in the first week.

NVIDIA is ready with a new high-end graphics card that will be slotted between the GeForce GTX 980 and the GeForce GTX TITAN-X, in its product stack. The GeForce GTX 980 Ti, as it's being called, will launch within the next couple of weeks, and will be based on the company's GM200 silicon. The core-configuration of this chip remains unknown, but it is rumored to feature 6 GB of GDDR5 memory, half that of the GTX TITAN-X, across its 384-bit wide memory bus.

Back in March, it was reported that NVIDIA will launch this SKU only after Summer. An early June launch suggests that either NVIDIA is spooked about AMD's Radeon R9 390X and its fancy-named sibling, which are expected to launch in the third week of June, and wants to siphon off high-end GPU sales as gamers and enthusiasts wrap up their Summer builds/upgrades, to settle down for the season's big game launches; or the company is confident of its performance, and really wants the GTX 980 Ti to appear on performance graphs, in reviews of AMD's new products.

Weeks ago, we were the first to report that AMD could give its top SKU carved out of the upcoming "Fiji" silicon a fancy name à la GeForce GTX TITAN, breaking away from the R9 3xx mold. A new report by SweClockers confirms that AMD is carving out at least one flagship SKU based on the "Fiji" silicon, which will be given a fancy brand name. The said product will go head on against NVIDIA's GeForce GTX TITAN-X. We know that AMD is preparing two SKUs out of the fully-loaded "Fiji" silicon - an air-cooled variant with 4 GB of memory; and a liquid-cooled one, with up to 8 GB of memory. We even got a glimpse of the what this card could look like. AMD is expected to unveil its "Fiji" based high-end graphics card at E3 (mid-June, 2015), with a product launch a week after.

AMD's upcoming Radeon R9 390X graphics cards will ship in two SKUs - an air-cooled one, with a moderately long reference design board (though not as long as the R9 290X), and a new Water-Cooled Edition (WCE) SKU, which will feature a very compact PCB - one that could be no bigger than that of the GeForce GTX 970 reference. This is possible because of AMD's HBM implementation. The 8 GB of memory on this card is present on the GPU package, as bare 3D-stacked DRAM dies, surrounding the GPU die, with an IHS covering everything; rather than the GPU package being surrounded by memory chips. Below is a mock-up of the card by ChipHell. It's not a picture. The radiator is off-proportions, the Radeon logo is misaligned, and the PCIe I/O is misaligned, etc. It should still give you a good idea of what the card looks like, particularly its length. Other specs on hand so far, include 4,096 GCN 1.2 stream processors, 256 TMUs, 128 ROPs, and a 4096-bit wide HBM interface, which at 1.25 GHz memory clock, will offer memory bandwidth of 640 GB/s.

While Fiji package will be bigger than that of, say, "Hawaii," overall the setup is more space-efficient, and conserves PCB real-estate. The PCB hence only has the GPU package and the VRM. AMD is doing away with the DVI connector on its reference PCB. It will only feature three DisplayPort 1.2a and one HDMI 2.0a. The WCE variant will feature a pump+block covering the GPU package, which will come factory-fitted to a 120 x 120 mm radiator. The air-cooled R9 390X will be longer, but only to house a heatsink and lateral blower. The single-GPU card could offer performance comparable to the dual-GPU R9 295X2, which is faster than the GeForce GTX TITAN-X. AMD CEO Lisa Su, speaking at the Investor Day event, in New York, on 6th May, hinted that the product could launch on the sidelines of either Computex 2015 (early-June) or E3 (mid-June).Image Courtesy: ChipHell. Many Thanks to GhostRyder for the tip.

NVIDIA reportedly adjusted launch of the GeForce GTX 980 Ti, to after Summer, 2015. The company could be using the time to let the market digest existing inventories of the GTX 980 (and avoid the repeat of slow sales on its GTX 770), particularly in the early-Summer season, when PC enthusiasts and gamers tend to upgrade or build afresh. The time could also be spent to watch what AMD comes up with, for its Radeon R9 390X.

The R9 390X, scheduled for a June-July launch, is based on a silicon that looks competitive with the GM200 on paper, and introduces a few new features, such as high-bandwidth memory (HBM). The GTX 980 Ti could feature an identical core-configuration to the GTX TITAN X, but feature half the memory amount at 6 GB, different clock speeds, and freedom for add-in card (AIC) partners to innovate custom-design cards.

NVIDIA is preparing its second GeForce graphics card based on its 28 nm GM200 silicon, which powers the $1,000 GTX TITAN-X. There are several rumors surrounding what NVIDIA could name the card. Some sources suggest NVIDIA could name it the GeForce GTX 980 Ti, while others point at names such as the GTX 990 (to set it sufficiently apart from the smaller GM204-based GTX 980).

The SweClockers report that sides with GTX 980 Ti for the name, mentions that the card could feature the chip's full complement of 3,072 CUDA cores, but feature 6 GB of memory, compared to 12 GB on the GTX TITAN-X. The memory bus width will stay at 384-bit. NVIDIA could allow its add-in card (AIC) partners to come up with custom-design cards, and so we could expect some cards with meaty cooling solutions (that keep the chip away from its 84°C temperature-throttle), and factory-overclocked speeds, to make the GTX 980 Ti even faster than the GTX TITAN-X. NVIDIA could time its launch with AMD's launch of the Radeon R9 390X.

Earlier today, AMD reportedly showed its industry partners (likely add-in board partners) a presentation, which was leaked to the web as photographs, and look reasonably legitimate, at first glance. If these numbers of AMD's upcoming flagship product, the Radeon R9 390X WCE (water-cooled edition) hold up, then it could spell trouble for NVIDIA and its GeForce GTX TITAN X. To begin with, the slides confirm that the R9 390X will feature 4,096 stream processors, based on a more refined version of Graphics CoreNext architecture. The core ticks at speeds of up to 1050 MHz. The R9 390X could sell in two variants, an air-cooled one with tamed speeds, and a WCE (water-cooled edition) variant, which comes with an AIO liquid-cooling solution, which lets it throw everything else out of the window in psychotic and murderous pursuit of performance.

It's the memory, where AMD appears to be an early adopter (as its HD 4870 was the first to run the faster GDDR5). The R9 390X features a 4096-bit wide HBM memory bus, holding up to 8 GB of memory. The memory is clocked at 1.25 GHz. The actual memory bandwidth will yet end up much higher than the 5.00 GHz 512-bit GDDR5 on the R9 290X. Power connectors will be the same combination as the previous generation (6-pin + 8-pin). What does this all boil down to? A claimed single-precision floating point performance figure of 8.6 TFLOP/s. Wonder how NVIDIA's GM200 compares to that. AMD claims that the R9 390X will be 50-60% faster than the R9 290X, and we're talking about benchmarks such as Battlefield 4 and FarCry 4. The expectations on NVIDIA's upcoming product are only bound to get higher.