Samsung Galaxy S3 GPU Benchmark Leaks; Absolutely Trumps Its Competition!

As the Galaxy S III unveiling date is coming closer, so are the rumors getting thick. Today’s rumor is about the GPU performance of the Galaxy S III.

The folks over at Phone Arena found the benchmark scores of the Galaxy S III GPU on GLBenchmark left by an unknown tester. The benchmark scores of the handset absolutely blow its competition out of the water. This includes NVidia’s Tegra 3 and Qualcomm’s S4, two of the biggest competition to the next Exynos SoC. The handset even managed to beat the iPhone 4S PowerVR SGX543MP2 GPU in quite a few handsets as well.

It is still unknown as to what GPU the SGS3 will use, but rumors indicate the Exynos 4412 SoC will pack in the same ARM Mali-400MP GPU as found in the Galaxy S2, except that it will be clocked higher. This higher clock speed will lead to about a 50% performance boost, on paper at least! Considering the Galaxy S III used in the benchmarks is still a prototype unit, there is a very high possibility that the retail device will perform even better!

Readers interested in checking out the benchmark scores should head over to Phone Arena for the appropriate screenshots.

 

AMD HD 7970 Officially Unveiled; Its A Beast In Every Sense!

AMD today launched its next-generation high-end graphics card, the Radeon HD 7970. The HD 7970 replaces the HD 6970, and is a beast in every word. The card packs in a whopping 4.3 billion transistors, and comes with 3GB of VRAM memory. The core of the 7970 is clocked at 925MHz, and comes with 2048 stream processors with a 384-bit memory bandwith bus.

The AMD HD 7970 is also the first GPU in the world to support the PCIE 3.0 slot and be compatible with DirectX 11.1. Thanks to the new ZeroCore power feature introduced by AMD in the HD 7970, the idle power draw of the GPU while displaying anything on the screen is a mere 15W, and less than 3W with a black screen. The maximum power consumption of the card is stated at 250W, just like its predecessor.

The company officials also state that the HD 7970 has a huge overclocking  potential, and can be easily overclocked to 1GHz+ speeds. Sadly, today’s announcement from AMD was just a soft-launch. The AMD HD 7970 won’t hit the retail stores in the United States before January 9th for a suggested retail price of $549. Interested buyers can read an in-depth review of the AMD HD 7970 over at AnandTech.

(Source)

AMD Fusion APUs Finally Unveiled

Today, at the ongoing CES 2011 AMD has announced their first ever family of Accelerated Processing Units (APU). The AMD Fusion based APUs will take on Intel’s Atom processors which are targeted towards netbook. According to AMD, their Fusion based APU’s provide a multi-core processor, a powerful Dx11 based GPU along with high definition (1080p) video playback and an all-day’ battery life.

As of now, AMD is offering two series of Fusion APUs E and C. The former is targeted primarily towards mainstream laptops, All-in-ones (AIO) and other small form factor based computers like HTPC. The latter is designed for HD netbooks and other emerging factors.

AMD_Fusion_APU

The E series APUs have a power consumption of 18W with clock speeds in the range of 1.5GHz to 1.6GHz. The C series APUs have a TDP of 9W with clock speeds of 1GHz and 1.2GHz. Both the series offer a dual-core as well as a single-core processor. AMD will also release A series of APUs in the first half of 2011 which will be targeted towards mainstream laptops. The A series APUs will pack up to 4 cores and a powerful Dx11 based discrete GPU.

AMD had to push back the release date of Fusion APUs numerous times before they finally unveiled it. Hopefully, the delays will be worth it and these Fusion APUs from AMD will provide stellar performance.

(Source)

Convert AMD HD 6950 To HD 6970 With A BIOS Flash

The hardware crazy folks over at Tech Power Up have brought a gift for all AMD HD 6950 owners out there. The gift is definitely going to make every AMD 6950 owner jump with joy.   The TPU (Tech Power Up) guys have found that the AMD HD 6950 can be converted to an AMD HD 6970 by a simple BIOS flash.

clip_image001

AMD launched the Cayman core based HD 6950 and the HD 6970 in the middle of this month. The main difference between the HD 6970 and the HD 6950 are the lower number of shaders and the slower core and memory clock speed of the latter. AMD HD 6950 owners just need to download the AMD HD 6970 BIOS and flash it on their HD 6950 GPU.

Until now, the TPU guys have managed to convert the HD 6950 to the HD 6970 from all popular GPU manufacturers including XFX, Sapphire and PowerColor. AMD HD 6950 owners can read the guide on how to convert their GPU to the HD 6970 here.

(Source)

AMD Unveils The Radeon HD 6950 And HD 6970 Graphic Cards

Today, AMD has unveiled its next generation high-end graphics card the HD 6970 and HD 6950. Both these cards are the first to be based on the new Cayman architecture. The core of the HD 6950 runs at 800MHz and packs in 1408 stream processors along with 88 texture units and 32 ROPs. The 2GB GDDR5 memory of the GPU is running at 1.25GHz.

AMD_HD_6970

The elder sibling of the HD 6950 the HD 6970 packs in 1536 stream processors, 96 texture units and 32 ROPs. The core of the GPU is clocked at 880 MHz while the 2GB GDDR5 memory runs at 1.375 GHz. Both the cards have a 256-bit memory bus width. The cards also support AMD’s EyeInfinity technology.

The HD 6950 and the HD 6970 go on sale starting from today from third parties like Power Color, Sapphire and XFX. AMD has priced both these cards very competitively at only $299 and $369 for the HD 6950 and the HD 6970 respectively.

Chromium Introduces GPU Rendering, Makes Chrome Better at Rendering Heavy Pages

Recently, a lot of work is being done to introduce GPU rendering into Google Chrome. New additions into web technologies like WebGL and 3D CSS has brought up a need for the web browser to be more CPU intensive. Most definitely, this is undesirable and Google Chrome is taking the right step in offloading the responsibility of rendering these components to the GPU. This will considerably improve the performance and responsiveness of Google Chrome.
google-chrome-logo

Google Chrome will feature a new GPU process in future versions that will manage all graphics related responsibilities. The GPU process will take in all graphics rendering tasks from the renderer process and send it to OpenGL or Direct3D. This access was not available to the renderer process earlier and neither is it now. However, the GPU process in question is allowed to run in a sandbox and have access to these graphics components of the OS.

This feature will be available for color conversion and scaling of videos. This will lighten the overburdened renderer process and give Google Chrome a smooth performance.

The idea of GPU rendering has just been implemented and the Chromium team wants to develop and advance more in this matter.

(Source)

Samsung Captivate Overclocked to 1.2Ghz

The Samsung Galaxy S is one of the best Android phones in the market right now. The phone has a 1Ghz Cortex A8 based Hummingbird processor, and 512MB of RAM. The Galaxy S features the best combination of CPU + GPUSamsung Captivate on an Android device. The Galaxy S is sold in the U.S under different operators and under various names like Captivate and Fascinate. With the various lag fixes available for it, the Galaxy S beats all other devices in Quadrant benchmark irrespective of their Android OS.

However, for some developers this was not enough and they have gone ahead and overclocked their Galaxy S to 1.2 GHz. The Hummingbird CPU inside the Galaxy S can be overclocked to 1.6 GHz safely. Interested readers need to flash a modified kernel to overclock their CPU to 1.2 GHz. They also need to download SetCPU from the Android Market. Here is the link on how to overclock Samasung Captivate to 1.2Ghz .

Keep in mind that the above link is valid only for Samsung Captivate owners. Samsung Galaxy S owners need to check this thread on how to overclock their CPU.

NVIDIA Announces DirectX 11 GeForce GTX 460 Graphic Card For $199

NVIDIA has announced a DirectX 11 GeForce GTX Graphic card which will sell for $199. The DX11 GeForce GTX 460 card comes in two versions, one with 768MB memory and the second one with 1GB memory.

NVIDIA GeForce GTX 460 GPU

NVIDIA today announced the latest addition to its Fermi-class of graphics processing units (GPUs), the NVIDIA ® GeForce GTX 460 — which delivers a great DX11 gaming experience and is priced right at the market’s $199 sweet spot.

The new GeForce GTX card is built from ground up and deliver up to 4x tessellation performance(i) over competing GPUs. The GTX 460 bring the ultimate gaming experience with DX11 bundled with and . GeForce GTX 460 also supports NVIDIA 3D Vision technology.

Epic Games, Tim Sweeney who is the Chairman, CEO and Technical Director commented: "With the GeForce GTX 460, NVIDIA is really upping the ante with an amazingly fast DirectX 11 GPU available at a great, mainstream-friendly price. Its new hardware tessellation support will enable games to reach an entirely new caliber of graphics."

GeForce GTX 460 Specifications and Pricing

Both the GTX 460 cards will have 336 CUDA Cores and 675 MHz Graphics Clock. GTX 460 will also have a 1350 MHz processor and a texture fill rate of 37.8 billion per second. The 768MB version will have a 192-bit memory interface and Memory Bandwidth of 86.4 GB per second.

On the other hand, the 1GB card will have a 256-bit Memory Interface Width and a Memory Bandwidth of 115.2 GB per second. The 768MB GeForce GTX 460 GPU will sell for $199, whereas the 1GB GPU will sell for $229.

The GeForce GTX 460 768MB version will be available starting today from NVIDIA add-in card partners. Wider availability of the GeForce GTX 460 1GB is expected by July 26, 2010.  For more information and specifications visit the product page for GeForce GTX 360.

Sony’s New PS3 Slim Model To Have A 40NM GPU

The Playstation 3 from Sony is arguably the best console – hardware wise – in the market currently. The PS3 has a Nvidia RSX GPU with 256MB of GDDR3 memory and, a Cell Microprocessor designed by IBM.PS3 RSX GPU With the launch of the PS3 Slim Model, the Cell CPU of PS3 moved to a 45nm manufacturing process. This led to a decrease in power consumption, lower temperatures, and reduction in the cost of production.

Now Playstation University reports that the new PS3 Slim RSX GPU will be based on the 40nm fabrication process.  At the time of its launch, the PS3 RSX GPU used the 65nm fabrication process. The shift to a lower fabrication process has many advantages including the reduction of the Yellow Light Of Death on the PS3s. Thanks to the shift in lower fabrication process, the cooling unit of the new PS3 Slim has also been redesigned. It is now less complex compared to its predecessor.

It is reported that the drop in power consumption is around 15%. The power supply has also seen a weight reduction, and now weighs in at 412g. The original PS3 power adapter used to weigh 815g. Quite a few chips have also been removed from around the Nvidia RSX GPU, leading to a lower cost of production.

The main benefit of the shift to 40nm fabrication process is the lower cost of production. The new PS3 Slim, even though cheaper by 100$, will allow Sony to earn profit on every console sold. Until now, Sony has been incurring losses on every PS3 they have sold.