4 ms·

I'm so interested to see what non-ML pipelines end up being possible & useful. Can we cobble these things into sound DSP processors? Or image convolvers? I'm no

I'm so interested to see what non-ML pipelines end up being possible & useful. Can we cobble these things into sound DSP processors? Or image convolvers? I'm not sure what architectural constraints there would be that would limit things down beyond what llvm can compile. > Note that these accelerators include an array of processors, while the LLVM backend only supports a single processor. Support for devices as a whole is available in open source tools based on MLIR Would be neat to see what interop with something like IREE looks like. Sure seems like MLIR is leading the rest-of-industry's (non-NV) efforts.

almostgotcaught 2y ago
> Would be neat to see what interop with something like IREE looks like https://github.com/nod-ai/iree-amd-aie https://github.com/nod-ai/iree-amd-aie
Archit3ch 2y ago
> Can we cobble these things into sound DSP processors? Not any more than the FPGAs they are based on. You need hard floating point for audio. I bet FFTs will be much faster on this hardware, though...
- Neywiny 2y ago
  The AI engines have floating point support. The AIE-ML emulates it. I don't know which kind these are based off of. The DSP58 of the Versal (which has the AI engines) also has floating point
imtringued 2y ago
It might be possible to make video encoders run on these NPUs. If I remember correctly the 8700G had 32 tiles. Even if the vectorization performance is weak, this should still encode a single 1080p stream at no CPU usage for the main CPU.