Add capability to use different allocators in cuML Python benchmarks #6903
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
csadorf
reviewed
Jun 12, 2025
| "--rmm-allocator", | ||
| choices=["cuda", "managed", "prefetched"], | ||
| default="cuda", | ||
| help="RMM memory resource to use (default: CUDA)", |
There was a problem hiding this comment.
Suggested change
| help="RMM memory resource to use (default: CUDA)", | |
| help="RMM memory resource to use (default: cuda)", |
Comment on lines
+226
to
+239
| # Setup RMM allocator based on command line option | ||
| if args.rmm_allocator == "cuda": | ||
| dev_resource = rmm.mr.CudaMemoryResource() | ||
| rmm.mr.set_current_device_resource(dev_resource) | ||
| print("Using CUDA Memory Resource...") | ||
| elif args.rmm_allocator == "managed": | ||
| managed_resource = rmm.mr.ManagedMemoryResource() | ||
| rmm.mr.set_current_device_resource(managed_resource) | ||
| print("Using Managed Memory Resource...") | ||
| elif args.rmm_allocator == "prefetched": | ||
| upstream_mr = rmm.mr.ManagedMemoryResource() | ||
| prefetch_mr = rmm.mr.PrefetchResourceAdaptor(upstream_mr) | ||
| rmm.mr.set_current_device_resource(prefetch_mr) | ||
| print("Using Prefetched Managed Memory Resource...") |
There was a problem hiding this comment.
I think this is a good use-case for match/case syntax:
Suggested change
| # Setup RMM allocator based on command line option | |
| if args.rmm_allocator == "cuda": | |
| dev_resource = rmm.mr.CudaMemoryResource() | |
| rmm.mr.set_current_device_resource(dev_resource) | |
| print("Using CUDA Memory Resource...") | |
| elif args.rmm_allocator == "managed": | |
| managed_resource = rmm.mr.ManagedMemoryResource() | |
| rmm.mr.set_current_device_resource(managed_resource) | |
| print("Using Managed Memory Resource...") | |
| elif args.rmm_allocator == "prefetched": | |
| upstream_mr = rmm.mr.ManagedMemoryResource() | |
| prefetch_mr = rmm.mr.PrefetchResourceAdaptor(upstream_mr) | |
| rmm.mr.set_current_device_resource(prefetch_mr) | |
| print("Using Prefetched Managed Memory Resource...") | |
| # Setup RMM allocator based on command line option | |
| match args.rmm_allocator: | |
| case "cuda": | |
| dev_resource = rmm.mr.CudaMemoryResource() | |
| rmm.mr.set_current_device_resource(dev_resource) | |
| print("Using CUDA Memory Resource...") | |
| case "managed": | |
| managed_resource = rmm.mr.ManagedMemoryResource() | |
| rmm.mr.set_current_device_resource(managed_resource) | |
| print("Using Managed Memory Resource...") | |
| case "prefetched": | |
| upstream_mr = rmm.mr.ManagedMemoryResource() | |
| prefetch_mr = rmm.mr.PrefetchResourceAdaptor(upstream_mr) | |
| rmm.mr.set_current_device_resource(prefetch_mr) | |
| print("Using Prefetched Managed Memory Resource...") | |
| case _: | |
| raise ValueError(f"Unknown RMM allocator type: {args.rmm_allocator}") |
Comment on lines
+223
to
+224
| # Define RMM allocator options | ||
| RMM_ALLOCATOR_TYPES = ["cuda", "managed", "prefetched"] |
There was a problem hiding this comment.
We don't use this variable anywhere, so I'd suggest to remove it.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR adds a new CLI option to run_benchmarks.py to set up different RMM allocators.