At a supercomputer conference in Germany, Intel unveiled plans for its new chip, expected to clock in at roughly 600 times the speeds of current processors, aiming at a position held by supercomputing rival Nvidia.
Six companies will show systems running a prototype version of Intel's Many Integrated Core architecture (MIC) at this week's International Supercomputer Conference.
The new chip, Intel contends, will be capable of ExaFLOP/s -- a quintillion computer operations per second -- hundreds times more than today's fastest supercomputers.
But don't expect this in your next home computer soon. The first prototypes are geared towards governments and universities, and aimed at feeding the computing needs of super computers.
MIC represents Intel's implementation of a graphics-based processor targeting use as a general-purpose accelerator for highly parallel applications.
The chip grew out of the company's Larrabee GPU effort and was relegated to a development platform chip in mid-2009.
It will compete with AMD's Radeon chips using OpenCL and Nvidia's graphics chips using its proprietary CUDA environment.
So far, Nvidia has a lead in this space which supports relatively few users buying chips at generally high prices and profit margins. For example, of 19 of the Top 500 supercomputers in the latest rankings that use graphics accelerators, 12 of them use Nvidia chips.
The current chip, called Knights Ferry, is not intended as a commercial product, in part because it does not support double precision floating point operations.
A future version, called Knights Corner and made in Intel's 22nm tri-gate process will support double precision, but Intel is not yet announcing when it will ship.
But it does hold the advantage of supporting the same programming models that most other Intel chips support, making adoption easier.
The programming model advantage of Intel's MIC architecture enabled us to quickly scale our applications running on Intel Xeon processors to the Knights Ferry Software Development Platform, said Professor Arndt Bode of the Leibniz Supercomputing Centre.
This workload was originally developed and optimized for Intel Xeon processors, but due to the familiarity of the programming model, we could optimize the code for the Intel MIC architecture within hours and also achieved over 650 GFLOPS of performance.
Knights Corner is planned for release as a product in 2012.