NVIDIA's New 28 nm Chip GeForce GTX680 Based On Kelper Architecture

The general issues faced when you go to buy a GPU are: power efficiency, performance, heat generated and the noise. Well, NVIDIA has answers for all these problems with a single chip. The California based company today launched new chips as GeForce GTX680 which is based on its next-generation architecture known as “Kelper”.

#-Link-Snipped-#

The size of the new chips is just 28 nm (almost billionth of a meter). The kelper architecture succeeds Fermi architecture which used to provide 40 nm chips. According to NVIDIA the new GPU is going to be the most power efficient and powerful GPU ever built. It has eight blocks for 1536 graphic cores as compared to 512 cores in Fermi architecture. The new GPU is extremely power efficient as it draws just 195 watts of power, way lower than GTX 580 which requires 250 watts of power. Less power draw makes it more power efficient and also helps in making less noise. According to Justin Walker, senior GeForce product manager, all cores of GTX680 share one single clock which makes it more power efficient. The basic clock runs at 1006 MHz and with “GPU boost”, you can boost it up to 1058 MHz.

According to NVIDIA, GeForce GTX680 is more than 300 percent efficient in DirectX 11 tessellation performance along with 43 percent performance boost and 23 percent less power consumption. There are two different versions, one for desktop and other for notebooks. For desktop gaming GeForce GTX680 GPU provides a faster and smoother experience while for notebooks GeForce 600M GPU does the trick.  Taiwan Semiconductor Manufacturer Company (TSMC) is going to build these chips for NVIDIA. The new chips will compete with AMD’s Southern Island chips which are also based on 28 nm architecture. The gamers can start experiencing high performance gaming for $499.