3 ms·

This is not true. Lots of algorithms simply can't use 100% of the GPU even though they're written as optimal as possible. FFT is one.

defrost 2y ago
In remote sensing | computation physicas applications it's rare to have a single FFT to compute (whatever algorithm is chosen). Hence the practice of stuffing many FFT's through GPU grids in parallel and working to max out the hardware usage in order to increase application throughput. eg: https://arxiv.org/pdf/1707.07263 https://arxiv.org/pdf/1707.07263 https://ieeexplore.ieee.org/document/9835388 https://ieeexplore.ieee.org/document/9835388
- shaklee3 2y ago
  I don't mean a single fft. I mean the fft algorithms are inherently not going to use the GPU at 100% utilization by any metric.
  - mpreda 2y ago
    Not so inherently IMO. What I mean is: where did you take that from? I program FFTs on GPUs, and I see no reason for the "inherently can't reach 100% utilization by any metric".
    - lights0123 2y ago
      I interpret that comment as you're not going to be using every silicon block that the GPU provides, like video codecs and rasterizing. If you've maxed out compute without going over the power budget, for example, you'd likely still be able to decode video if the GPU has a separate block for it.
      - defrost 2y ago
        I had a similar read .. I packed a lot of parallel FFT's and other processing into custom TI DSP cards but the DSP family chips were RISC and carried little 'baggage' - just fat fat 32 bit | 64 bit floating point pipelines with instruction sets optimised for modular ring indexing of scalar | vector operations. Even then they ran @ 80% "by design" for expected hard real time usage .. they only went to 11 and dropped results in toast until they smoke tests and with operators that redlined limits (and got feedback to that effect).
    - shaklee3 2y ago
      I'd be curious to see how you can do it. Try launching an fft of any size and batches and see if you can hit 100%

defrost 2y ago

In remote sensing | computation physicas applications it's rare to have a single FFT to compute (whatever algorithm is chosen). Hence the practice of stuffing many FFT's through GPU grids in parallel and working to max out the hardware usage in order to increase application throughput. eg: https://arxiv.org/pdf/1707.07263 https://arxiv.org/pdf/1707.07263 https://ieeexplore.ieee.org/document/9835388 https://ieeexplore.ieee.org/document/9835388

shaklee3 2y ago
I don't mean a single fft. I mean the fft algorithms are inherently not going to use the GPU at 100% utilization by any metric.
- mpreda 2y ago
  Not so inherently IMO. What I mean is: where did you take that from? I program FFTs on GPUs, and I see no reason for the "inherently can't reach 100% utilization by any metric".
  - lights0123 2y ago
    I interpret that comment as you're not going to be using every silicon block that the GPU provides, like video codecs and rasterizing. If you've maxed out compute without going over the power budget, for example, you'd likely still be able to decode video if the GPU has a separate block for it.
    - defrost 2y ago
      I had a similar read .. I packed a lot of parallel FFT's and other processing into custom TI DSP cards but the DSP family chips were RISC and carried little 'baggage' - just fat fat 32 bit | 64 bit floating point pipelines with instruction sets optimised for modular ring indexing of scalar | vector operations. Even then they ran @ 80% "by design" for expected hard real time usage .. they only went to 11 and dropped results in toast until they smoke tests and with operators that redlined limits (and got feedback to that effect).
  - shaklee3 2y ago
    I'd be curious to see how you can do it. Try launching an fft of any size and batches and see if you can hit 100%