Discussion on tuning machinery

[alazzaro]

Follow up of #804

[RMeli]

• The performance gain with the tuned A100 kernels is minor compared to using the P100 kernels like the tuned P100 kernels work reasonably well for V100.

• It is better to use the full set of autotuned and predicted kernels from the previous GPU generation than to use only a relative small set of autotuned kernels.

From the comments above (see #804 (comment)) looks like

Then, the strategy will be to rename the file/parameters in "AMD" and "NVIDIA" and drop the specific GPU version. As I said, I will add a generic kernel which will be good enough for all cases we don't cover with autotuning.

is a good compromise to move forward. But I'm no expert on this, so it's good to hear what people think about this issue.

[hfp]

Then, the strategy will be to rename the file/parameters in "AMD" and "NVIDIA" and drop the specific GPU version.

Good idea. In particular since a specific tuning may also need maintenance given the underlying runtime version can change over time (aka new CUDA version). Also, this opens a reasonable option to tune/refresh for the latest/deployed GPU (and to naturally phase-out some tuning for older GPUs, not saying it would not run anymore).