GeForce GTX 680 Nvidia's new graphics

The new Nvidia graphics processors are more powerful than the previous generation, consume less energy and have automatic overclocking.

 More power, automatic overclocking and less energy expenditure. These were the promises of Nvidia in San Francisco, where the company unveiled the secrets of the new architecture Kepler reporters.

 GTX 680 first graphics cards to hit the Kepler market as always the top of the range. These include 1536 graphics core running at 1006 MHz, a speed that can automatically go up to 1058 MHz through the functionality of automatic overclocking the GPU Boost. The GPU boost constantly analyzes various factors such as temperature and energy. If there is room for maneuver increases the processing speed automatically without any user intervention.

The graphic Kepler also include 2 GB of GDDR5 RAM with 256-bit bandwidth and 6 Gbps capacity and compatibility with PCIE 3.0

Kepler is an improved architecture compared to the previous generation with 3.5 billion transistors. According to Nvidia, much more power than the generation of graphics cards to replace. This increased power results in a complete redesign of the graphics. For example, a GTX 580 had groups each processing core 32 and a controller. By comparison, each group of the GTX 680 has 192 processors, six times more than the previous generation. Eight of these groups are the nuclei that make 1536 such a GTX 680.

Besides the hardware, the new GTX 680 support new antialiasing modes. According to Nvidia, the TXAA provides eight times quality of MSAA and costing the same in terms of performance of the MSAA 4x.

Finally, we Adaptive VSync, which allows us to enable VSync without the huge losses associated performance.

The average sale price suggested by Nvidia for the GTX 680 is 419 euros, excluding VAT. With VAT, the price should be € 520 in Portugal. However, we must bear in mind that these are the prices suggested by Nvidia and not by the partners who actually sell the graphics cards.

The Informatics Review is already testing the GeForce GTX 680.

Read More →

Review iPad 3 Performance

Apple says the iPad 3  "four times faster than the chip Tegra 3", competitor Nvidia. If the statement is true, many wonder. First test results provide answers, at least partial.

Vietnamese site Tinhte published the first benchmark results, which are not at the level of Apple's claims. Using the Geekbench benchmark that measures the performance of floating point operations, a test of memory and processing stream reveals that "brute force", which offers 1.0 GHz ARMv7 A5X of dual-core CPU remained the same from previous generations . At the same time, it is confirmed definitively that the new iPad 3 comes with 1 GB of internal memory RAM, which is double what is already installed.

What then is left to allegations of Apple? The graphics and GPU core is what distinguishes the new creation of Apple's competition. GPU is actually variant formerly known SGX543MP2 PowerVR chips, which comes under the label SGX543MP4. To make this chip is known to charge Imagination Technologies, a specialist in the field of graphics for notebooks. The theoretical maximum that provides SGX543MP4 MPolygon 134/4 is GPixel / s fill rate.

You should also know that the tests were conducted by well-known site that AnandTech showed that the past generation iPad Tablet PC was faster than 30 to 80 percent of the Tegra 3 graphics in certain situations. If Apple kept its promise to accelerate the existing core, one would expect that the graphics performance to some extent be the order of the promise of "four times better performance than the Tegra 3."

Note here also that the above is a "clock watch" comparison, since Tegra actually clocked higher speed, the difference is further reduced by a higher frequency of Nvidia chip.

Read More →