[cuda] Dump useful GPU characteristics #13955
Merged
Conversation
This file contains 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
antiagainst
requested review from
benvanik,
GMNGeoffrey and
stellaraccident
as code owners
This commit implements `iree_hal_cuda_driver_dump_device_info` to print out GPU characteristics including launch configuration size limits, per block/multiprocessor resource limits, memory system characteristics, and others. On NVIDIA GeForce RTX 2070 SUPER, `iree-run-module --dump_devices` shows: ``` - gpu-compute-capability: 7.5 - driver-max-cuda-version: 12.1 - launch-max-block-dims: (1024, 1024, 64) - launch-max-grid-dims: (2147483647, 65535, 65535) - block-max-thread-count: 1024 - block-max-32-bit-register-count: 65536 - block-max-shared-memory: 49152 bytes - multiprocessor-max-thread-count: 1024 - multiprocessor-max-block-count: 16 - multiprocessor-max-32-bit-register-count: 65536 - multiprocessor-max-shared-memory: 65536 bytes - memory-has-unified-address-space: 1 - memory-supports-managed-memory: 1 - memory-can-map-host-memory-to-device: 1 - memory-supports-pageable-memory-access-from-device: 0 - memory-supports-concurrent-managed-access: 1 - memory-supports-memory-pools: 1 - memory-l2-cache-size: 4194304 bytes - gpu-multiprocessor-count: 40 - gpu-clock-rate: 1815000 kHz - gpu-warp-size: 32 - kernel-has-execution-timeout: 1 ```
antiagainst
force-pushed
the
cuda2-attributes
branch
from
423b016
to
928d70d
Compare
benvanik
approved these changes
Jun 5, 2023
NatashaKnk
pushed a commit
to NatashaKnk/iree
that referenced
this pull request
Jul 6, 2023
This commit implements `iree_hal_cuda_driver_dump_device_info` to print out GPU characteristics including launch configuration size limits, per block/multiprocessor resource limits, memory system characteristics, and others. On NVIDIA GeForce RTX 2070 SUPER, `iree-run-module --dump_devices` shows: ``` - gpu-compute-capability: 7.5 - driver-max-cuda-version: 12.1 - launch-max-block-dims: (1024, 1024, 64) - launch-max-grid-dims: (2147483647, 65535, 65535) - block-max-thread-count: 1024 - block-max-32-bit-register-count: 65536 - block-max-shared-memory: 49152 bytes - multiprocessor-max-thread-count: 1024 - multiprocessor-max-block-count: 16 - multiprocessor-max-32-bit-register-count: 65536 - multiprocessor-max-shared-memory: 65536 bytes - memory-has-unified-address-space: 1 - memory-supports-managed-memory: 1 - memory-can-map-host-memory-to-device: 1 - memory-supports-pageable-memory-access-from-device: 0 - memory-supports-concurrent-managed-access: 1 - memory-supports-memory-pools: 1 - memory-l2-cache-size: 4194304 bytes - gpu-multiprocessor-count: 40 - gpu-clock-rate: 1815000 kHz - gpu-warp-size: 32 - kernel-has-execution-timeout: 1 ```
nhasabni
pushed a commit
to plaidml/iree
that referenced
this pull request
Aug 24, 2023
This commit implements `iree_hal_cuda_driver_dump_device_info` to print out GPU characteristics including launch configuration size limits, per block/multiprocessor resource limits, memory system characteristics, and others. On NVIDIA GeForce RTX 2070 SUPER, `iree-run-module --dump_devices` shows: ``` - gpu-compute-capability: 7.5 - driver-max-cuda-version: 12.1 - launch-max-block-dims: (1024, 1024, 64) - launch-max-grid-dims: (2147483647, 65535, 65535) - block-max-thread-count: 1024 - block-max-32-bit-register-count: 65536 - block-max-shared-memory: 49152 bytes - multiprocessor-max-thread-count: 1024 - multiprocessor-max-block-count: 16 - multiprocessor-max-32-bit-register-count: 65536 - multiprocessor-max-shared-memory: 65536 bytes - memory-has-unified-address-space: 1 - memory-supports-managed-memory: 1 - memory-can-map-host-memory-to-device: 1 - memory-supports-pageable-memory-access-from-device: 0 - memory-supports-concurrent-managed-access: 1 - memory-supports-memory-pools: 1 - memory-l2-cache-size: 4194304 bytes - gpu-multiprocessor-count: 40 - gpu-clock-rate: 1815000 kHz - gpu-warp-size: 32 - kernel-has-execution-timeout: 1 ```
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.
This commit implements
iree_hal_cuda_driver_dump_device_infoto print out GPU characteristics including launch configuration size limits, per block/multiprocessor resource limits, memory system characteristics, and others.On NVIDIA GeForce RTX 2070 SUPER,
iree-run-module --dump_devicesshows:Progress towards #13245