Slashdot Mirror

← Back to Stories (view on slashdot.org)

High performance FFT on GPUs

Posted by ryuzaki0 on from the testing-it-out dept.

A reader writes: "The UNC GAMMA group has recently released a high performance FFT library which can handle large 1-D FFTs. According to their webpage, the FFT library is able to achieve 4x higher computational performance on a $500 NVIDIA 7900 GPU than optimized Intel Math Kernel FFT routines running on high-end Intel and AMD CPUs costing $1500-$2000. The library is supported for both Linux and Windows platforms and is tested to work on many programmable GPUs. There is also a link to download the library freely for non-commerical use."

3 of 274 comments (clear)

  1. Rush hour math. by Anonymous Coward · · Score: 3, Insightful

    ""The UNC GAMMA group has recently released a high performance FFT library which can handle large 1-D FFTs. According to their webpage, the FFT library is able to achieve 4x higher computational performance on a $500 NVIDIA 7900 GPU than optimized Intel Math Kernel FFT routines running on high-end Intel and AMD CPUs costing $1500-$2000. "

    GPUs are nice, but there's the little matter of getting data and results on and off the chip.

  2. $1500-$2000? by Wesley+Felter · · Score: 3, Insightful

    I sense a little bias here; the fastest Intel and AMD processors are actually $1,000.

  3. Re:It's nice... by edp · · Score: 3, Insightful
    "If you are taking data off of some kind of sensor, there are damned few sensors with 24 good bits of data out of the noise floor. Radars work just fine with 16-bit A/D converters."

    Take a look at their benchmarks. The chart goes up to eight million elements. The accumulated rounding error in FFT outputs may be around n * log2(n) ULP, where n is the number of elements, and ULP (units in last place) is relative to the largest input element. (Caveats: That is the maximum; the distribution of the logs of the errors resembles a normal distribution. Input was numbers selected from a uniform distribution over [0, 1). The error varies slightly depending on whether you have fused multiply-add and other factors.)

    So with eight million elements, the error may be 184 million ULP, or over 27 bits. With only 24 bits in your floating-point format, that is a problem. Whether you had 24-bit or 1-bit data to start with, it is essentially gone in some output elements. Most errors are less than the maximum, but it seems there is a lot of noise and not so much signal.

    It may be that the most interesting output elements are the ones with the highest magnitude. (The FFT is used to find the dominant frequencies in a signal.) If so, those output elements may be large relative to the error, so there could be useful results. However, anybody using such a large FFT with single-precision floating-point should analyze the error in their application.