Skip to content

[CI][Benchmarks] update to latest compute-benchmarks #18267

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 5 commits into from
May 2, 2025
Merged

[CI][Benchmarks] update to latest compute-benchmarks #18267

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
2a1d272
Update compute.py
reble Apr 30, 2025
57be7d8
Renaming NoEvents to UseEvents and swap value
reble Apr 30, 2025
f620b23
Adding explicit option
reble Apr 30, 2025
79ef8fd
Enable GraphApiSubmitGraph benchmark for all runtimes
reble Apr 30, 2025
b1dc639
Merge branch 'intel:sycl' into pablo/update-ci-benchmarks
reble May 1, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 5 additions & 4 deletions devops/scripts/benchmarks/benches/compute.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ def git_url(self) -> str:
return "https://github.com/intel/compute-benchmarks.git"

def git_hash(self) -> str:
return "9c1ed6fd59a7a40f8829251df4b5c0d847591183"
return "5e4e39cabd0f74422754890ff8e362789c714543"

def setup(self):
if options.sycl is None:
Expand Down Expand Up @@ -168,7 +168,7 @@ def benchmarks(self) -> list[Benchmark]:
benches.append(UllsKernelSwitch(self, runtime, 8, 200, 0, 0, 1, 1))

# Add GraphApiSubmitGraph benchmarks
for runtime in self.enabled_runtimes([RUNTIMES.SYCL, RUNTIMES.UR]):
for runtime in self.enabled_runtimes():
for in_order_queue in [0, 1]:
for num_kernels in [4, 10, 32]:
for measure_completion_time in [0, 1]:
Expand Down Expand Up @@ -640,7 +640,7 @@ def bin_args(self) -> list[str]:


# TODO: once L0 SubmitGraph exists, this needs to be cleaned up split benchmarks into more groups,
# set all the parameters (NoEvents 0/1, which should get inverted into UseEvents) and
# set all the parameters (UseEvents 0/1) and
# unify the benchmark naming scheme with SubmitKernel.
class GraphApiSubmitGraph(ComputeBenchmark):
def __init__(
Expand Down Expand Up @@ -681,7 +681,8 @@ def bin_args(self) -> list[str]:
f"--InOrderQueue={self.inOrderQueue}",
"--Profiling=0",
"--KernelExecutionTime=1",
"--NoEvents=1", # not all implementations support NoEvents=0
"--UseEvents=0", # not all implementations support UseEvents=1
"--UseExplicit=0",
]


Expand Down
Loading