If you're weren't lucky enough to attend 2013's Consumer Electronics Show, then you can catch up on NVIDIA's keynote here. The company had a fairly talked about showing, revealing a similarly praised and criticized product in its Project Shield, Android-based gaming console. NVIDIA also revealed its new Tegra 4 processor, all of which you can see in its entirety in the video after the break.
Read More | Twitch TV
It is being reported that OS X 10.8.3 beta has ATI Tahiti 7000 drivers support and can be found in the system profiler. This could very well mean that Mountain Lion is ready for an imminent Mac Pro refresh. In addition, the new iMacs are said to have Nvidia graphics, so this support isn't mean for those--no current Macs has these graphic chips, so it's good news for those hardcore users that need extra muscle and have been holding out.
Read More | Netkas
The perpetual war for supremacy between AMD and Nvidia constantly leaves enthusiasts dodging shrapnel: When you want the best video card you can afford, why buy one now instead of waiting for the better one the competing chipset designer will undoubtedly release in a few months? This leaves reviewers in a tough spot, too, as we're constantly proclaiming that nearly every new card is the fastest ever. But because you can only live in the world you live in, we're obliged to go there. So, here goes once again: The just-released AMD Radeon HD 7970 ($549 list) is the latest fastest and most feature-rich single-GPU card ever, surpassing our previous Editors' Choice winner, the Nvidia GeForce GTX 580.
Though we have little doubt that Nvidia will yank back that title with its next generation of cards, the 7970 is an impressive achievement for the moment. (It's rumored to become widely available by early January of 2012.) The inaugural member of the Southern Islands family, it utilizes a fresh architecture AMD refers to as "Graphics Core Next." Based on a new 28nm process technology and utilizing more than 4.3 billion transistors, Graphics Core Next uses a revised instruction set architecture, gives each compute unit the ability to simultaneously execute instructions from multiple kernels, and delivers an increased number of instructions per clock cycle per square millimeter of GPU space. The result, so AMD claims, is "designed for high utilization, high throughput, and multitasking."
The Radeon HD 6790 is positioned between the upper-end-mainstream Radeon HD 6770 and the lower-end-enthusiast Radeon HD 6850, intended for users running at midrange resolutions (such as 1,680-by-1,050) and detail settings. The card offers 1.34 teraflops of compute power; has a core clock speed of 840 MHz, 800 stream processors, 40 texture units, and 16 ROPs; and is loaded with 1GB of GDDR5 frame buffer, operating at 4.2 Gbps on a 256-bit memory path.
AMD's press materials for the 6790 identify the 6790's TDP as about 150 watts—more than the 6850, and approximately in the same league as the card above that one, the Radeon HD 6870 (151 watts). AMD's reference spec for the card suggests it will require two six-pin connectors from a power supply, though the company says that some models will be available using only one.
The months-long jockeying for position between AMD and Nvidia has led to this moment: Who has the faster flagship video card? Nvidia held the crown for a long while thanks to its powerful and polished GTX 580, still the best single-processor card on the market. But when AMD released its dual-GPU Radeon HD 6990 earlier this month, and it delivered blistering benchmark results along with a sky-high $699 list price and an ultra-noisy fan, it looked like AMD might own the top tier this generation. Now that Nvidia has released its own dual-GPU card, the Nvidia GeForce GTX 590 (also $699), we definitively know the answer: AMD just wins the performance crown. Nvidia's card has some solid reasons to recommend it—much better noise characteristics, it will fit in a (slightly) wider variety of cases—but for this much money you probably want the fastest card there is. And the GTX 590, in spite of its virtues, is not quite it.
The GTX 590 is, however, packed with power. You'd expect that from any two-GPU card in general—the last one Nvidia released was the GTX 295, in early 2009—and especially from one that essentially fuses two powerful GF110 GPUs (the kind used in the GTX 580). It sports a total of 1,024 CUDA processing cores, 128 texture units, 96 ROP units, and 32 tessellation engines for making the most of one of the most sought-after DirectX 11 (DX11) features. The card's graphics clock runs at 607 MHz, its processor clock at 1,215 MHz, and its memory clock at 3,414 MHz. It's loaded with 3,072MB of GDDR5 memory for the frame buffer, which operates over a 384-bit memory interface.
Hot on the heels of AMD's recently released Radeon HD 6990, Nvidia introduced its own new flagship video card today, the GeForce GTX 590.
Nvidia's first dual-GPU video card since the GTX 295 in early 2009, the GTX 590 unites a pair of GF110 GPUs (the kind used in the GTX 580, the fastest single-GPU card on the market) on a single card. This means you get of 1,024 CUDA processing cores, 128 texture units, 96 ROP units, and 32 tessellation engines. The card's graphics clock runs at 607 MHz, its processor clock at 1,215 MHz, and its memory clock at 3,414 MHz. It's loaded with 3,072MB of GDDR5 memory for the frame buffer, which operates over a 384-bit memory interface.
For gamers and enthusiasts who aren't willing to settle for just an ordinary video card—you know, the kind with only one GPU—the Radeon HD 6990, which AMD is releasing today, could be the next must-have product.
As the 6000-series replacement for 2009's ATI Radeon HD 5970, the 6990 boasts a pair of powerful graphics processing units (GPUs), and some of the speediest specs on the market: compute power of 5.1 teraflops, a core clock speed of 830 MHz, 3,072 stream processors, 192 texture units, 64 ROPs, and 4GB of GDDR5 frame buffer running a long a 5-Gbps, 256-bit memory path. For adventurous users who want even faster frame rates, a switch on the card itself can automatically overclock the card to still-higher levels of performance.
Newegg has a deal on the PNY GeForce GTS 250 video card. This one has 1GB of DDR3 RAM, supports HDCP, and supports SLI. If you pick up two of them, that means they’ll play nicely together for an even nicer performance boost. You can pick up the PNY GeForce GTS 250 for $119.99, a good $30 off the original price. Take a look at the PNY GeForce GTS 250 on Newegg to get the savings.
As always, you can find all sorts of Newegg promo codes and deals on our forums.
Read More | PNY GeForce GTS 250
We take a few moments with AMD‘s Rick Bergman, General Manager of the Graphics Processing Group, about their new Spider platform—a three-tiered computing system that helps tie together and ensure compatibility between the chipset, CPU and GPU. We talk about what sort of benefits this means to the enthusiast-class consumer. We also dive into the nitty-gritty with the new Radeon, currently codenamed the R870—a dual chip design that has two separate GPU units with an internal Crossfire link and Alternate-Frame Rendering to tie them together. I also ask about some of the design challenges realized in smashing two cards together and getting an efficient, workable result. The design is impressive and the ability to link two of these cards together will come to mean a great deal of graphics processing power in the near future.
The card will be available late January and can be powered in a typical system by a 500 watt power supply; ATI has cut their power requirements for similarly powered cards in half, without sacrificing any performance.