We are always a little reluctant to post gossip like this, yet there seems to be some sort of foundation for this rumor. It appears that more information on GT300, the new GeForce/Quadro/Tesla flagship graphics card chip is on the loose. First of all, the GPU will have 512 Shader processors, and that's a big step up from the current 240 Shader processors.
Next to that, NVIDIA supposedly is moving from SIMD to MIMD architecture. Long story short: SIMD always has the advantage in raw peak operations per second. After all, it mainly consists of as many adders, floating-point units, shaders.
SIMD (Single Instruction, Multiple Data) refers to a parallel computer that runs the exact same program, that’s the “single instruction” part – on each of its simultaneously-executing parallel units. MIMD can pack more of their stuff onto chips, too, using the same density.
MIMD (Multiple Instruction, Multiple Data) refers to parallel computing that runs an independent separate program – that’s the “multiple instruction” part – on each of its simultaneously-executing parallel units.
Why NVIDIA makes this move is a little unclear, but basically the GPU becomes a huge floating point unit, which is great for games of course .. yet even better for GPGPU activity and CUDA related applications.
The new GPU will have 16 clusters of 32 shaders. Though it is unknown how fast they will be clocked, rumors state that if they can be clocked at 2 GHz, we'd be nearing a cool 3 teraflops of compute performance (right now we are hovering around one teraflop) per GPU. GT300 would give direct hardware access [HAL] for CUDA 3.0, DirectX 11, OpenGL 3.1 and OpenCL.
There is no release date on GT300 but we do know it'll be somewhere this year. But from the looks of it, that's a big can of you know what that NVIDIA will be delivering:
Site
Navigation
