NVIDIA today unveiled the NVIDIA Jetson TX2, a credit card-sized platform that delivers AI computing at the edge -- opening the door to powerfully intelligent factory robots, commercial drones and smart cameras for AI cities. Jetson TX2 offers twice the performance of its predecessor, or it can run at more than twice the power efficiency, while drawing less than 7.5 watts of power. This allows Jetson TX2 to run larger, deeper neural networks on edge devices. The result: smarter devices with higher accuracy and faster response times for tasks like image classification, navigation and speech recognition.

"Jetson TX2 brings powerful AI capabilities at the edge, making possible a new class of intelligent machines," said Deepu Talla, vice president and general manager of the Tegra business at NVIDIA. "These devices will enable intelligent video analytics that keep our cities smarter and safer, new kinds of robots that optimize manufacturing, and new collaboration that makes long-distance work more efficient."

Feast your eyes on the first image of a reference NVIDIA GeForce GTX 1080 Ti graphics card, compared side by side with the PCB of the company's flagship (still) TITAN X Pascal. As you can see, the GTX 1080 Ti is based on the same PCB as the TITAN X Pascal, since the two cards are based on the same "GP102" chip (albeit with different core configurations). To begin with, the GTX 1080 Ti features 11 memory chips, compared to 12 on the TITAN X Pascal, on account of its narrower 352-bit GDDR5X memory interface. It makes up for the narrower memory bus with faster 11 Gbps memory chips, than the 10 Gbps chips found on the TITAN X Pascal.

The main difference between the GTX 1080 Ti and TITAN X Pascal, however, is NVIDIA bolstering the VRM with a 2x dual-FET design. NVIDIA basically placed an additional set of MOSFETs and capacitors along all the blank traces of the reference PCB. This approach lowers the load on each individual MOSFET, in turn lowering VRM temperatures. It probably also enables a higher power-limit. NVIDIA also updated the reference design cooling solution with a new vapor-chamber base-plate. The cooler also exhausts through the entire width of the second slot in the card's I/O shield. This meant sacrificing the DVI connector. The GeForce GTX 1080 Ti goes on sale later this month, priced at US $699.

INNO3D, a leading manufacturer of high-end hardware components and computer utilities, introduces its new flagship INNO3D GeForce GTX 1080 Ti graphics cards. The new range will be available in Founder Edition and the premium iChiLL series. Designed on NVIDIA's Pascal architecture, INNO3D GeForce GTX 1080 Ti iChiLL X3 and X4 simply own the concept of performance tops the previous flagship by no less than 35%. This makes it even faster in games than the world-famous TITAN X also built on NVIDIA's architecture. The new hero GPU comes packed with extreme gaming horsepower, utilizing next-gen 11 Gbps GDDR5X memory, and a massive 11GB frame buffer.

The new GeForce GTX 1080 Ti iChiLL X3 and X4 GPUs also come equipped with INNO3D's unmatched iChiLL cooling solution -awarded for better performance at lower noise levels- improves the overall active temperature control. The Herculez Armor back-plate protects the cards from outer damages as well and puts it way at the top of Ultra High End graphics cards in the world.

NVIDIA is reportedly working on a major driver update that improves performance across the board on games that take advantage of DirectX 12. The company is working with major game studios to improve DirectX 12 performance by up to 16 percent on recent AAA titles, such as "Gears of War 4," "Hitman" (2016), "Rise of the Tomb Raider" and "Tom Clancy's The Division." Even "Ashes of the Singularity" is getting a performance improvement.

At its GeForce GTX 1080 Ti launch event, NVIDIA revealed that its "Pascal" and "Maxwell" GPU architectures are capable of tile-based rendering, a feature that significantly improves video memory bandwidth utilization. With the GTX 1080 Ti going on sale later this month, one can expect NVIDIA to release new drivers.

Looking back on NVIDIA's GDC presentation, perhaps one of the most interesting aspects approached was the implementation of tile-based rendering on NVIDIA's post-Maxwell architectures. This has been an adaptation of typically mobile approaches to graphics rendering which keeps their specific needs for power efficiency in mind - and if you'll "member", "Maxwell" was NVIDIA's first graphics architecture publicly touted for its "mobile first" design.

This approach essentially divides the screen into tiles, and then rasterizes the entire frame in a per-tile basis. 16×16 and 32×32 pixels are the usual tile sizes, but both Maxwell and Pascal can dynamically assess the required tile size for each frame, changing it on-the-fly as needed and according to the complexity of the scene. This looks to ensure that the processed data has a much smaller footprint than that of the full image rendering - small enough that it makes it possible for NVIDIA to keep the data in a much smaller amount of memory (essentially, the L2 memory), dynamically filling and flushing the available cache as possible until the full frame has been rendered. This means that the GPU doesn't have to access larger, slower memory pools as much, which primarily reduces the load on the VRAM subsystem (increasing available VRAM for other tasks), whilst simultaneously accelerating rendering speed. At the same time, a tile-based approach also lends itself pretty well to the nature of GPUs - these are easily parallelized operations, with the GPU being able to tackle many independent tiles simultaneously, depending on the available resources.

NVIDIA today unveiled the GeForce GTX 1080 Ti graphics card, its fastest consumer graphics card based on the "Pascal" GPU architecture, and which is positioned to be more affordable than the flagship TITAN X Pascal, at USD $699, with market availability from the first week of March, 2017. Based on the same "GP102" silicon as the TITAN X Pascal, the GTX 1080 Ti is slightly cut-down. While it features the same 3,584 CUDA cores as the TITAN X Pascal, the memory amount is now lower, at 11 GB, over a slightly narrower 352-bit wide GDDR5X memory interface. This translates to 11 memory chips on the card. On the bright side, NVIDIA is using newer memory chips than the one it deployed on the TITAN X Pascal, which run at 11 GHz (GDDR5X-effective), so the memory bandwidth is 484 GB/s.

Besides the narrower 352-bit memory bus, the ROP count is lowered to 88 (from 96 on the TITAN X Pascal), while the TMU count is unchanged from 224. The GPU core is clocked at a boost frequency of up to 1.60 GHz, with the ability to overclock beyond the 2.00 GHz mark. It gets better: the GTX 1080 Ti features certain memory advancements not found on other "Pascal" based graphics cards: a newer memory chip and optimized memory interface, that's running at 11 Gbps. NVIDIA's Tiled Rendering Technology has also been finally announced publicly; a feature NVIDIA has been hiding from its consumers since the GeForce "Maxwell" architecture, it is one of the secret sauces that enable NVIDIA's lead.

In the wake of its $699 GeForce GTX 1080 Ti enthusiast-segment graphics card launch, NVIDIA lowered the pricing of its previous GeForce "Pascal" flagship, the GeForce GTX 1080, which is now down to USD $499, from its launch price of $599 (custom-design baseline pricing). The GTX 1070 is unflinched for now, from its $349 baseline pricing. This launch should turn up the heat on AMD, with its HBM-powered Radeon RX Vega pricing, much in the same way the GTX 980 Ti steered pricing of the Radeon R9 Fury X.

NVIDIA today released a hotfix version of its GeForce driver suite. Numbered the 378.77, these fix some issues with the previous driver release, numbered 378.66. Solved issues include driver installation errors for laptops with GeForce GTX 1050 and 1050 Ti GPUs, crashes in Minecraft and some other Java-based titles, and resolving 'Debug Mode' as default option on Pascal based GPUs.

This driver release contains all previous improvements, including an optimized experience for Sniper Elite 4, For Honor and Halo Wars 2, whilst also introducing support for NVIDIA's Video SDK 8.0, which provides support for high-bit-depth (10/12-bit) hardware accelerated video decoding of VP9 and HEVC. You can, as always, grab these hot-from-the-coding-gods drivers from the link below, right here on TPU.DOWNLOAD: NVIDIA GeForce 378.77 Hotfix Drivers

NVIDIA's GeForce.com website today was updated to greet us with a curious-looking countdown to evening, February 28. The countdown goes with the caption "It's Almost Time," with "Ti" in bold lettering. This just about confirms launch of the GeForce GTX 1080 Ti, NVIDIA's next high-end graphics card based on the same "GP102" silicon as the TITAN X Pascal. It remains to be seen if the company endows the SKU with more CUDA cores than the TITAN X Pascal, or less. A lot will depend on what NVIDIA's product managers learned about AMD's upcoming Radeon Vega.

NVIDIA could attempt to steal the limelight from AMD's 2017 "Capsaicin & Cream" launch event for its Ryzen desktop processors, slated for February 28, with a parallel GeForce GTX event along the sidelines of the 2017 Game Developers' Conference (GDC). At this event, the company is expected to launch its next enthusiast-segment graphics card, the GeForce GTX 1080 Ti. This could at least be a paper-launch, with market availability following through in March.

While the GTX 1080 Ti is a graphics card, and Ryzen a processor (they don't compete), NVIDIA's choice of launch-date could certainly steal some attention away from AMD's big day. Besides launching Ryzen, it wouldn't surprise us if AMD teases its upcoming Radeon "Vega" graphics cards a little more. The GeForce GTX 1080 Ti is expected to be based on the same "GP102" silicon as the company's flagship TITAN X Pascal graphics card, and could be positioned very close to the USD $1,000 mark, given that NVIDIA priced the TITAN X Pascal at a wallet-scorching $1,199.

NVIDIA (NASDAQ: NVDA) today reported revenue for the fourth quarter ended January 29, 2017, of $2.17 billion, up 55 percent from $1.40 billion a year earlier, and up 8 percent from $2.00 billion in the previous quarter. GAAP earnings per diluted share for the quarter were $0.99, up 183 percent from $0.35 a year ago and up 19 percent from $0.83 in the previous quarter. Non-GAAP earnings per diluted share were $1.13, up 117 percent from $0.52 a year earlier and up 20 percent from $0.94 in the previous quarter.

For fiscal 2017, revenue reached a record $6.91 billion, up 38 percent from $5.01 billion a year earlier. GAAP earnings per diluted share were $2.57, up 138 percent from $1.08 a year earlier. Non-GAAP earnings per diluted share were $3.06, up 83 percent from $1.67 a year earlier. "We had a great finish to a record year, with continued strong growth across all our businesses," said Jen-Hsun Huang, founder and chief executive officer of NVIDIA. "Our GPU computing platform is enjoying rapid adoption in artificial intelligence, cloud computing, gaming, and autonomous vehicles.‎

NVIDIA today introduced a range of Quadro products, all based on its Pascal architecture, that transform desktop workstations into supercomputers with breakthrough capabilities for professional workflows across many industries. Workflows in design, engineering and other areas are evolving rapidly to meet the exponential growth in data size and complexity that comes with photorealism, virtual reality and deep learning technologies. To tap into these opportunities, the new NVIDIA Quadro Pascal-based lineup provides an enterprise-grade visual computing platform that streamlines design and simulation workflows with up to twice the performance of the previous generation, and ultra-fast memory.

"Professional workflows are now infused with artificial intelligence, virtual reality and photorealism, creating new challenges for our most demanding users," said Bob Pette, vice president of Professional Visualization at NVIDIA. "Our new Quadro lineup provides the graphics and compute performance required to address these challenges. And, by unifying compute and design, the Quadro GP100 transforms the average desktop workstation with the power of a supercomputer."

Ashes of the Singularity seems to be the benchmark tool of choice for upcoming AMD products, for some reason; and it was once again used to benchmark an upcoming AMD Ryzen processor. The benchmark results were quickly deleted after they were posted, but the hardware enthusiast should never be underestimated, and timely screenshot skills always help keep alive these little slips of the trade.

Unlike some previous benchmark leaks of Ryzen processors, which carried the prefix ES (Engineering Sample), this one carried the ZD Prefix, and the last characters on its string name are the most interesting to us: F4 stands for the silicon revision, while the 40_36 stands for the processor's Turbo and stock speeds respectively (4.0 GHz and 3.6 GHz). This is the 8-core, 16-thread SMT-enabled monster of a processor that AMD will be bringing to the table in its uphill battle against Intel, with the Ryzen chip having achieved CPU Framerate scores of 81.4 (normal batch, 73.4 (medium batch) and 60.2 (heavy batch), paired with a Pascal-based NVIDIA Titan X (which would likely point towards the test having been done by an independent, off-AMD labs part).

NVIDIA released the GeForce 378.57 Hotfix drivers, designed to address glaring bugs. The driver update fixes a critical game crash bug noticed in "Minecraft" and other Java-based titles; and resolves "debug mode" being the default mode on GPUs based on the "Pascal" architecture. The rest of their features appear consistent with older releases. Find the driver from the links below.DOWNLOAD: NVIDIA GeForce 378.57 Hotfix

After disappointing scores of potential buyers by skipping a GTX 1080 Ti announcement at CES - which could have been a last-moment decision on the company's part when AMD failed to make any relevant VEGA announcement - it looks like NVIDIA has chosen the grand stage of PAX East, which begins at March 10th, as the place to carry the previously-confirmed addition to their Pascal line of GPUs.

This information (which should be taken with a maybe unhealthy grain of salt) came to light by way of an MSI (NVIDIA's AIB partner) representative, which also mentioned that the 1080 Ti would be available from board partners (including, naturally, MSI itself) at time of launch.

With a TDP of 180W, it was only a matter of time before we started to see ever more compact GeForce GTX 1080 graphics cards, since it is after all based on the third largest implementation of the "Pascal" architecture (after GP100 and GP102). The ZOTAC GeForce GTX 1080 Mini is the smallest GTX 1080 implementation to date, with its length of 8.3" inches (just about 21cm), and its PCB even shorter than that. ZOTAC used the difference in lengths of the cooler and PCB to cram in the longer and wider portions of the aluminium fin-stacks.

The cooler of the ZOTAC GeForce GTX 1080 Mini features five 6 mm-thick copper heat pipes, which pass through a dense aluminium fin-stack, which is then ventilated by 90 mm (left) and 100 mm (right) fans. The fan in the front appears to be of a quieter, low-RPM type, while the one on the rear end is a high air-flow fan. A neat back-plate covers up the reverse side of the PCB. An LED-lit ZOTAC logo is located on the top. The card ships with factory-overclocked speeds of 1620 MHz core, 1753 MHz GPU Boost, and an untouched 10 GHz (GDDR5X-effective) memory, compared to reference speeds of 1607/1733 MHz. The card draws power from a single 8-pin PCIe power connector. Display outputs include three DisplayPorts and one each of HDMI and DVI. This design should allow ZOTAC to achieve two market positions - charge a premium for the compact size; or sell at a lower price than reference, if the competition heats up.

At CES 2017, Digital Storm made sure to showcase the computer case to devour all others: the Aventum 3 stands tall and deep with 28" height and more than 25" front-to-back, this is a truly massive tower, with room for more fans (it has about 20), cables (more than can be sanely counted), and custom liquid cooling than two or three mid-tower-size typical desktops.

BIOSTAR is proud to announce its latest addition to its line of gaming products with its new VGA for gamers looking for the perfect balance of price and performance with the new BIOSTAR GeForce GTX 1050 Ti. With an MSRP of $159, this card will be the perfect upgrade gift for yourself or a gamer friend. Enjoy gaming to the next level with its new dual-fan cooling for the best performance the GTX 1050 Ti can bring.

Want SLI to work right on your GeForce GTX 1080, GTX 1070, or TITAN X Pascal without having to spend $50 on a fancy SLI bridge with an LED light-show you don't need? Look no further than the "no-frills" NVIDIA SLI HB bridge by ASUS. The bridge is just a fiberglass PCB with SLI slots where you'd want them, and simple ASUS ROG branding printed on the PCB, nothing too fancy. The bridge comes with 60 mm or 2-slot spacing (1 slot between two dual-slot thick graphics cards). The bridge will be sold separately for a fraction of the price of the company's LED-lit bridges; as well as being part of some of the company's upcoming motherboards' accessory packages.

New reports have surfaced that indicate NVIDIA is all set to launch a new mid range mobile graphics solution, the GTX 1050 M at CES 2017 in Las Vegas. While NVIDIA has already released mobile variants of the GTX 1080, GTX 1070 and GTX 1060, people not willing to spend big on higher tier products from the green camp are limited to "Maxwell" based GTX 960 M or GTX 950 M offerings. Reports also indicate there has been somewhat of a surge in demand for GTX 1060 M equipped laptops, where the new GTX 1050 M could be nicely positioned to further augment consumer appetite.

As we reported in November, we can expect that in line with the existing "Pascal" based mobile solutions, the new GTX 1050 M to sport the same (or better) core-configuration as its desktop counterpart, but with significantly higher clock speeds. This should make the GTX 1050 M an attractive option as it would endow the laptop with graphical horsepower exceeding the very capable GTX 970 M. All in all with new Intel "Kaby Lake" processors set to take the stage at CES 2017 too, we could see quite an array of new or reconfigured laptops scattered throughout the show.

In what is surely the biggest GTX 1080 announcement this year (pun intended), Colorful, at their Colorful CGU (Colorful Games Union) event, has taken the lids of their KUDAN version of the GTX 1080 - a monstrous, 4-slot graphics card sure to vie for most of your rig's interior space. A hybrid design is used to keep the Pascal graphics processor cooled - it actually features a triple-slot, triple-fan over an aluminum heatsink array on top of the GPU and circuitry, as well as a water cooling-powered backplate, that can be used in conjunction with the card's own air cooling or your custom water-loop.

The rest is business as usual: 2560 CUDA cores and 8 GB GDDR5X VRAM, with the core being clocked at 1936 MHz. The memory is clocked at reference speeds of 10 GHz but the addition of an overclocking BIOS should help alleviate that fact. Power to the card is fed through dual 8-Pin connectors which should be enough to power the electrical circuitry on the PCB, which Colorful touts as having been fully re-engineered so as to deliver the KUDAN's required power characteristics. Display outputs on the card include 1x DVI, 1x HDMI and 3x DisplayPort.

No details on pricing, but expect this one to come in near the top of overall GTX 1080 offerings. The GTX 1080 KUDAN is expected to launch somewhere along December or January.

Jon Peddie Research (JPR), the graphics and multimedia industry's research and consulting firm, announced estimated PC graphics add-in-board (AIB) shipments and suppliers' market share for Q3'16. The last two quarters have seen both NVIDIA and AMD release and expand a new AIB line-up, the former launching the GeForce 10 series, powered by "Pascal" GPU's and the latter releasing the Radeon 400 series, powered by "Polaris" GPU's.

JPR's AIB Report tracks PC add-in graphics boards (AIBs), which carry discrete graphics chips. AIBs are used in desktop PCs, workstations, servers, and other devices such as scientific instruments. They can be factory installed or sold directly to customers as aftermarket products. In all cases, AIBs represent the higher end of the graphics industry using discrete chips and private high-speed memory, compared to the integrated GPUs in CPUs or SOCs that share slower system memory.

NVIDIA has today released the latest Hotfix GeForce drivers that address the low memory clocks on certain factory-overclocked "Pascal" cards. The driver includes all of the previous updates from the 375.86 WHQL driver with the "Pascal" memory clock fix listed as the only difference. You can Grab the drivers from the links below.DOWNLOAD: NVIDIA GeForce 375.95 Beta for Windows 10 64-bit | Windows 10 32-bit | Windows 8/7/Vista 64-bit | Windows 8/7/Vista 32-bit

If you have installed the GeForce 375.86 WHQL drivers you may experience a variety of issues. According to users on Guru3D forums, and indeed users here on our very own forums, the drivers released just yesterday have had reports where certain games frame rates are cut in half, however more alarmingly many users also report memory clock speed drops resulting in on-screen flickering and instability.

The memory speed and flickering issues appear to be limited to cards using the "Pascal" architecture, NVIDIA have acknowledged that the issues are widespread and work is already underway to remedy them. We expect either that a hotfix or new WHQL driver set will follow shortly, depending on their release cycle.

Air-cargo shipping manifest descriptions point to the possibility that NVIDIA's upcoming high-end graphics card based on the GP102 silicon, the GeForce GTX 1080 Ti, could feature 10 GB of memory, and not the previously thought of 12 GB, which would be the same amount as the TITAN X Pascal. NVIDIA apparently is getting 10 GB to work over a 384-bit wide memory interface, likely using chips of different densities. The GTX 1080 Ti is also rumored to feature 3,328 CUDA cores, 208 TMUs, and 96 ROPs.

NVIDIA has, in the past, used memory chips of different densities to achieve its desired memory amount targets over limited memory bus widths. The company achieved 2 GB of memory across a 192-bit wide GDDR5 memory interface on the GeForce GTX 660, for example. The product number for the new SKU, as mentioned in the shipping manifest, is "699-1G611-0010-000," which is different from the "699-1G611-0000-000" of the TITAN X Pascal, indicating that this is the second product based on the PG611 platform (the PCB on which NVIDIA built the TITAN X Pascal). NVIDIA is expected to launch the GeForce GTX 1080 Ti in early-2017.