AMD's new Radeon HD 6900 series occupies the top position in their single-GPU product hierarchy. The two models, the HD 6950 and HD 6970 are very much like the HD 5850 and HD 5870 that they replace. The xx50 cards generally run at a lower clock rate and have a few sections of the GPU disabled, presumably because the vendor is trying to reclaim chips that have a small, isolated manufacturing defect. But what happens when your manufacturing process is so good that you're not producing enough "defective" chips to meet the market demand? When is a 6950 not a 6950? Well, quite often, as it turns out. In the case of the PowerColor PCS++ Radeon HD 6950 video card, it just depends on which way you flip the switch. Push it one way and you have a standard Radeon HD 6950, with 1408 shaders running at 800 MHz. Push it the other way and you have 1536 shaders running at 880 MHz, which is the exact configuration of the HD 6970. The only difference is that PowerColor kept the 1250 MHz memory chips in the PCS++, instead of springing for the 1500 MHz memory, like a real HD 6970 has. Join Benchmark Reviews as we investigate this unique product from PowerColor, that's sure to cause some excitement in the marketplace.

The VisionTek Killer HD5770 Combo Video Card combines a Bigfoot Networks Killer E2100 NIC and a Radeon HD5770 GPU to offer a complete gaming solution that undercuts the price of buying both seperately, also combining the network controller with the video card takes more work off your CPU's shoulders allowing it to be better utilized in other areas of your games. I'm sure you are thinking that the HD5770 is probably not the most powerful GPU out there and that VisionTek could have maybe used something with a bit more grunt but this combination of technologies takes the original VisionTek HD5770 from 7.5 inches long to 10 inches long, so it makes sense for obvious reasons. Benchmark Reviews aims to provide you with an unbiased review of the VisionTek Killer HD5770 and report back our findings, keeping you informed on the latest technologies available on the market today.

It's hard to believe the last time I used a Titanium NVIDIA GPU was in 2002. The GeForce GTX 560 Ti is the new boom in the "mid-end" sector. It replaces the highly overclock-able GTX 460 video card and adds extra MHz and cores increasing 20-40% performance in many games. As it could be expected, brands like ASUS, GIGABYTE or MSI will pair this GPU with a special heatsink-cooler and increase factory frequencies to sell it at a higher price. Benchmark Reviews tests the ASUS ENGTX560 Ti DCII TOP. DCII means this card now comes with an improved version of the Direct CU cooler, now with 2 fans and adding dissipation area. They also tag this product as TOP because it sports a 900MHz GPU core instead of the 822MHz from the reference design and 1050MHz instead of 1000MHz from stock. Aside from that, ASUS also bundles their very own overclocking/monitoring software to modify frequency and voltage values and add some extra performance for free. Let's have a look at the ASUS ENGTX560 Ti and check if it can be a good contender against the competition, including the highly acclaimed GeForce GTX 460.

History always repeats itself, and NVIDIA is hoping for a repeat performance with their new GTX 560 Ti video card. With performance per mm² and per watt numbers that put the first Fermi chips to shame, the GTX 460 deserves all the success it has enjoyed. Now, they've applied the same tweaks to the GF104 GPU that brought the high-end Fermi cards back from the dead, and we have the new GF114 in place to do battle in the upper mid-market. MSI is also repeating the same steps they used with the GTX 460 to capture the interest of graphics overclocking enthusiasts. Their Twin Frozr II design offered a massive upgrade in performance compared to the stock cooling scheme, and it helped achieve a world record of 1.0 GHz on air cooling for the GTX 460. Fast forward six months and MSI has launched an upgraded GTX 560 Ti video card at the same time that the reference design is hitting the market. The MSI N560GTX-Ti Twin Frozr II/OC model offers up a modest 7% factory overclock, but I'm guessing there's more where that came from. Benchmark Reviews takes full advantage of MSI's hardware upgrade, as we explore the full potential of NVIDIA's latest midrange marvel on our test bed.

It's been nearly a decade since NVIDIA last used the Titanium moniker on one of their product, and for those who can still recall how the GeForce 4 series was revision of the previous series the new GeForce GTX 560 Ti will make perfect sense. Replacing the GeForce GTX 470 video card in the current product stack, NVIDIA's GeForce GTX 560 Ti uses a tuned GF114 GPU that finally delivers a full GF104 Fermi architecture. The original GF104 GPU offered seven of eight possible Streaming Multiprocessors (SM) with the GeForce GTX 460 video card, and now NVIDIA returns to enable that last SM to make even more cores available to GF114, now 384 compared to 336. Keeping with tradition, the GeForce GTX 560 Ti uses an identical SM configuration as the GeForce GTX 460. Each SM still offers 48 CUDA cores, four dispatch units, and eight texture/special function units. Besides including the eighth and final SM on the GPU, what's different is the myriad of transistor-level changes to improve power efficiency and in turn allowed for significantly faster clock speeds. In this article, Benchmark Reviews tests the EVGA GeForce GTX 560 Ti (model 01G-P3-1561-AR) against an entire market of graphics card options.

It's been nearly a decade since NVIDIA last used the Titanium moniker on one of their product, and for those who can still recall how the GeForce 4 series was revision of the previous series the new GeForce GTX 560 Ti will make perfect sense. Replacing the GeForce GTX 470 video card in the current product stack, NVIDIA's GeForce GTX 560 Ti uses a tuned GF114 GPU that finally delivers a full GF104 Fermi architecture. The original GF104 GPU offered seven of eight possible Streaming Multiprocessors (SM) with the GeForce GTX 460 video card, and now NVIDIA returns to enable that last SM to make even more cores available to GF114, now 384 compared to 336. Keeping with tradition, the GeForce GTX 560 Ti uses an identical SM configuration as the GeForce GTX 460. Each SM still offers 48 CUDA cores, four dispatch units, and eight texture/special function units. Besides including the eighth and final SM on the GPU, what's different is the myriad of transistor-level changes to improve power efficiency and in turn allowed for significantly faster clock speeds. In this article, Benchmark Reviews tests the GeForce GTX 560 Ti against an entire market of graphics card options...

When we talk about different video card brands, there's always a factor which motivates us to choose one over any other. Most likely, we make our decisions depending on retail price, but there are things to consider: the bundle and accessories, factory overclocked speeds, and of course, included heatsinks and fans so that the GPU can be overclocked higher or simply work without being as loud and hot as a reference design. With this in mind, Benchmark Reviews tests the GIGABYTE GV-R685OC-1GD AMD Radeon HD 6850 video card. We've already tested some HD 6850 GPUs before, but GIGABYTE offers their newest design with the Windforce 2x GPU cooler and Ultra Durable VGA technology. Additionally, this is the factory OC version which brings 820MHz (against 775MHz) GPU Core clock and 4200MHz (instead 4000MHz) GDDR5 Memory clocks. Let's analyze the GV-R685OC-1GD model and see if it can be a serious contender against reference HD 6850 and GTX 460 graphics cards.

In the wake of the HD 6900 series debut last month, the Barts GPU and its place in the product line finally makes perfect sense. The Cypress chip, as used on the Radeon HD 5830 with 334 square millimeters of silicon, is way too big for just 1120 shaders and 16 ROPS. The new Barts GPU uses just 255 mm² to do the same job only better, thanks to twice the number of ROPs as the 5830. Although TSMC denied AMD the opportunity to roll out 32nm-based chips for this product cycle, they went back to the drawing board and optimized this new generation of GPUs for the current 40nm manufacturing process. AMD did a better job of reading the rule book last time, and their first full line of 40nm GPUs dominated the graphics market for a full 6 months. AMD has now successfully inserted an additional class of GPU (as defined by die size), to fill the performance gap that existed in the Radeon HD 5000 series. You may have seen some benchmarks already for some reference HD 6870 cards, but please follow along with Benchmark Reviews as we take a complete look, inside and out, at the PowerColor PCS+ HD6870 1GB GDDR5 graphics card.