Vibhu Jawa
Vibhu Jawa
@nv-dlasalle , I agree with your approach too and actually lines up better with what we want to do with cuGraph. This was initially suggested in https://github.com/dmlc/dgl/issues/4326 but based on...
Lets close this issue. https://github.com/rapidsai/cugraph/pull/3269 fixed most things
This was closed by: https://github.com/rapidsai/rmm/pull/1168 , Can we close this ?
@beckernick , Could you able to fetch the logs from the workers (when you see this error again) , i suspect they might have some more context.
Related dask issue: https://github.com/dask/dask/issues/5809 Which has a example here: https://github.com/dask/dask/issues/5809#issuecomment-634291208
There are two things here we are discussing here. 1. Using RMM Pool by default This comes down to choosing perf over potential memory contention from other libraries . In...
> On 2, there's at least [a blogpost](https://medium.com/rapids-ai/pytorch-rapids-rmm-maximize-the-memory-efficiency-of-your-workflows-f475107ba4d4) now. Maybe we can do more to promote that? I agree that blog post is what I link so that is very...
@ChrisJar , Thanks for sharing these benchmarks. Do you have thoughts on how this can change if we scale to 10K. Not saying we should prioritize that, just wondering if...
CC: @ChrisJar for awareness.