Items tagged with GPU

There is one constant in the world of supercomputers: no one is going to be 'fastest' for long. In fact, some supercomputers can seem downright slow after only a couple of years, as hardware continues to become faster and more dense. Take for example Oak Ridge's TITAN supercomputer. Launched in 2013, this supercomputer managed to push about 20 petaFLOPs of throughput (17.59 pFLOPs LINPACK; 27 pFLOPs theoretical). At the time, that was downright mind-blowing. But consider TaihuLight, China's latest supercomputer, coming in at 93 petaFLOPs. For those who don't want to grab a calculator or exercise... Read more...
You’ve been hearing about AMD’s next-gen Polaris GPU architecture and the first card to feature it, the Radeon RX 480, for months now. We’ve already covered a number of Polaris’ new features and architectural enhancements, and have even shown you a few renders of the Radeon RX 480 itself, but now we’ve got a card in hand and are ready to get down to some in-house testing. Feast your eyes on this...The AMD Radeon RX 480 Has ArrivedAs you can see, the AMD Radeon RX 480 has a design language that’s reminiscent of the previous-gen Radeons, with an understated, rectangular fan shroud and dimpled faceplate.... Read more...
Building your own computer can be an extremely satisfying experience, but that doesn't mean that anyone has to shrug off potential improvements that would make that experience even smoother, right? Have you ever wanted to, for example, make the motherboard and graphics card into a single unit? If you didn't, the folks at Colorful did, and the result is one of the least colorful motherboards out there, the B150-GP104: To accommodate such a powerful graphics card into a motherboard design, Colorful had to make some obvious tweaks to the overall design. Gone are full-sized PCIe slots, for example.... Read more...
Raja Koduri, SVP of AMD RTG Shows Off World's First 14nm GPU- Radeon RX 480When AMD’s Radeon Technologies Group (RTG) first disclosed details of its upcoming Polaris GPU architecture, it called it a “historic leap in performance per watt for Radeon GPUs”. If you don’t follow the GPU scene closely, Polaris is AMD RTG’s next-gen graphics architecture that will be produced using both TSMC’s 16nm and Global Foundries 14nm FinFET manufacturing processes.The goal with Polaris is to help drive Radeon GPUs down to lower overall power envelopes and enable better performance per watt. Polaris-based GPUs... Read more...
The powerful GeForce GTX 1080 got the lion’s share of media attention immediately following NVIDIA CEO Jen-Hsun Huang’s official unveiling at an event in Austin, Texas a couple of weeks back. But its little brother, the GeForce GTX 1070, was also mentioned along with its expected price point, which just happened to be far more attainable for most PC enthusiasts. Better yet, the GeForce GTX 1070 would reportedly offer “Titan X class performance”, which is particularly impressive, given the 1070’s sub-$400 asking price of partner-built boards. Like the 1080, the GTX 1070 Founder's Edition... Read more...
A couple of weeks back, at a packed event in Austin, Texas filled with tech press and attendees from the DreamHack gaming event that took place up the road, NVIDIA CEO Jen Hsun-Huang unveiled the company’s newest graphics cards, the GeForce GTX 1080 and GeForce GTX 1070. These latest flagships represent a true generational leap in performance and efficiency, thanks to NVIDIA's new Pascal GPU architecture at their core and the bleeding-edge 16nm FinFET + VLSI manufacturing processes used to produce the chips. For the last few years, virtually all of the discrete GPUs being produced used... Read more...
A few weeks back, during its GPU Technology Conference, NVIDIA announced a new Quadro-branded graphics card for the professional workstation market. That card is the Quadro M2000 we’ll be showing you here today. The Quadro M2000 is a low-power pro graphics solution featuring a GPU based on NVIDIA’s Maxwell architecture. In fact, it’s the very same GPU used on one of the more mainstream GeForce GTX-branded cards. However, with the Quadro M2000, additional features and capabilities are enabled through NVIDIA's Quadro software/driver suite and the card’s particular hardware configuration.The Quadro... Read more...
AMD has officially turned on the hype machine for Polaris, its next generation GPU architecture that will eventually be the foundation of a new family of Radeon GPUs from top to bottom. In anticipation of its launch later this year, AMD has erected a dedicated Polaris page on its website that highlights the various technologies and features involved. The new microsite also provides what looks to be the first glimpse of a naked Polaris GPU. It appears in a banner at the top of the page—the GPU is seen floating in space surrounded by stars (Polaris is the brightest star in the constellation Ursa... Read more...
At the risk of not wanting to count chickens before they hatch, it's pretty hard to not be excited for AMD right now. For so many years, the company has struggled to increase its share value due to a lack of truly competitive products. However, this week, we've heard of two examples of how AMD could turn itself around, and the benefits are already materializing. A couple of days ago, we reported that AMD struck a deal to license its processor and SoC technologies so that Tianjin Haiguang Advanced Technology Investment Co., can create its own processors based on AMD's technologies. The deal is reported... Read more...
At last week’s GPU Technology Conference, NVIDIA’s CEO Jen-Hsun Huang unveiled a couple of key products and technologies that target the High-Performance Computing, or HPC, space, including the Tesla P100 data center accelerator and its companion DGX-1 deep-learning system, which is a powerful server that has up to eight Tesla P100 cards at its core. Today though, we have more information about the underlying architecture employed in the P100, otherwise known NVIDIA’s Pascal GPU architecture.Pascal is the follow-up to the Maxwell architecture, which is leveraged in NVIDIA’s current-generation... Read more...
AMD just released some details regarding its upcoming seventh generation of A-Series and FX APUs, formerly codenamed Bristol Ridge. Bristol Ridge is essentially a refresh of the previous generation Carrizo design and uses the same 28nm process. In lieu of a die shrink, however, AMD has made improvements to the manufacturing process, tweaked the transistor mix, and enabled new features that were previously dormant in Carrizo, and the end result is higher frequencies at a given power level, in addition to some IPC (instructions per clock) improvements. The newly activated features... Read more...
We're all eagerly awaiting AMD's next GPU architecture codenamed Polaris, which is being built on a 14-nanometer FinFET manufacturing process and is due out later this year. But if we look further ahead, there's something even more exciting in store. It's called "Project Greenland," otherwise known as "Vega," details of which were leaked to the web over the weekend.Yu Zheng, an R&D Manager at AMD Shanghai, listed Project Greenland as part of his work experience on his LinkedIn profile. The listing contained several interesting details that AMD probably didn't authorize Zheng to disclose, and... Read more...
Professional system builders are usually obsessed with silence and cooling. Custom cooling solutions are some of the biggest differentiators in pre-built systems. And one of the chief selling points of aftermarket cooling gear for enthusiasts is noise output, usually listed in decibels. There are even companies, like "be quiet!" for example, that focus on noise output as a main selling point for entire line-ups of cases, power supplies, and heat sinks. CPUs and GPUs are typically what get the most attention when it comes to aftermarket cooling.  Stock coolers on GPUs usually do an adequate... Read more...
Although the names, or more specifically the acronyms, for some of the new technologies NVIDIA is unveiling today leaked out recently, details on what they offered gamers and how they worked were scarce. At the Game Developers Conference currently underway in San Francisco, however, NVIDIA officially took the wraps of some new graphics techniques for shadows and lighting – dubbed Hybrid Frustum Traced Shadows (HFTS), NVIDIA Volumetric Lighting, and Voxel Accelerated Ambient Occlusion (VXAO) – which can drastically improve in-game image quality. And the features are already being leverages in a... Read more...
1 2 3 4 5 Next ... Last