CHIP DESIGNER Nvidia has launched its Kepler GPU with the Geforce GTX680 and Geforce GT640M for laptops.
Nvidia's Kepler GPU architecture is built using TSMC's 28nm process node and succeeds Fermi, which the firm launched two years ago. Given AMD's successful Radeon HD7900 series launch late last year, the heat was on Nvidia to put a competitive Kepler out the door as soon as possible, and now the firm has announced both desktop and laptop versions of its Kepler graphics chip.
Nvidia's Geforce GTX680 comes with 2GB of 256-bit GDDR5 memory and features 1,536 cores running at a base clock speed of 1,006MHz with a bump up to 1,058MHz under load. The card's TDP tops out at 195W and the card requires two six-pin power connectors.
While Nvidia's reference Geforce 680 'Boost' frequency is a mere 52MHz, board vendors such as EVGA have already set aside SKUs with 'to be decided' clock speeds, suggesting there is some headroom to up Kepler's clock speed. After all, with AMD's Southern Islands chips showing that TSMC's 28nm process node does leave some headroom, enthusiasts will want to know just how far they will be able to push Nvidia's 28nm TSMC fabbed chip.
Nvidia also announced its Geforce GT640M for laptops, however the real fireworks are expected to be set off once Intel lifts its embargo on laptops that make up its Ivy Bridge launch programme. Until then Nvidia's Geforce GT640M has up to 394 cores and runs at 625MHz with 2GB of GDDR3 or GDDR5 memory.
Nvidia's Geforce GTX680 graphics card is being priced in the £420 range, making it almost exactly the same price as AMD's ultra high-end Radeon HD7970. All eyes will be on AMD to see how it tries to one-up Nvidia again, with the shortest odds laid on a two-GPU vanity card to replace the Radeon HD6990. µ
Companies need to rate limit posts based on keywords, warns Trend Micro
Uses 20 percent less power than traditional systems
Sign up for INQbot – a weekly roundup of the best from the INQ