Code-named “Kepler,” the next major GPU architecture from Nvidia, is slated for 2011 release. Kepler will be based on a 28-nanometer fabrication process, and Nvidia intends for it to be nearly three times more efficient, in DP FLOPS per watt, than today’s Fermi architecture. Nvidia CEO Jen-Hsun Huang noted that such an improvement goes “far beyond” what process technology advances alone can achieve. Changes in the chip architecture, design, and software will contribute to that advance, as well.
And, the Maxwell architecture will come next, in 2013, and will be produced on a 22-nm fabrication process. Maxwell promises nearly an 8X increase in DP FLOPS per watt beyond today’s Fermi chips. Huang noted that, in parallel computing, power is the primary constraint, which’s why he chose that metric to describe future architectures.
Between now and the Maxwell generation, Nvidia plans to introduce several compute-focused features into its GPU architectures, including pre-emption.