You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
A Benchmark Result from Web Frameworks Benchmark(the-benchmarker)
#762
The rankings below were recalculated from the public results dataset included in the-benchmarker/web-frameworks PR #10356. The implementation is CWIST v4.3.1, while the results dataset labels its version as 4.3.
There is no official single overall ranking, so the tables show requests-per-second rankings at each concurrency level.
Concurrency
CWIST req/s
CWIST overall rank
CWIST GC req/s
CWIST GC overall rank
GC throughput difference
64
193,781.31
41/402
185,025.09
51/402
-4.52%
256
176,707.29
41/400
166,733.78
55/400
-5.64%
512
170,503.71
45/399
162,997.75
53/399
-4.40%
Average latency is 4.45%, 6.22%, and 4.80% higher in GC mode, respectively. At all three concurrency levels, both implementations report zero HTTP errors, timeouts, and socket connection/read/write errors.
CWIST download tag recorded in the results: v4.3.1
Benchmark host: AMD Ryzen 7 5700U, 16 CPUs, x86_64 Linux
Measurement method: GET /, zrk --closed, keep-alive, approximately 15 seconds, concurrency levels 64/256/512
GitHub Actions run 44035943561 for this PR provides no result artifacts. Its CWIST jobs perform only builds and 21 API contract tests, not load measurements. Accordingly, the figures below come from the complete public results dataset included in the PR, not from numbers in CI logs. The PR dataset is newer than the dataset on the base branch.
Rankings use every implementation with the same metric at the same concurrency level. Throughput is ranked from highest to lowest, while latency and standard deviation are ranked from lowest to highest. Each rank is calculated as 1 + the number of strictly better values. Some implementations have no results at concurrency levels 256 and 512, leaving denominators of 400 and 399, respectively.
Values have been rounded for readability. Because these are aggregated results, total requests and received bytes may also contain fractional values or rounding. Server CPU saturation is approximately 69–71%, so these measurements should not be described as fully saturated under the project's 75% guideline.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
the-benchmarker/web-frameworks#10356
CWIST / CWIST GC Benchmark Results
Summary
The rankings below were recalculated from the public results dataset included in
the-benchmarker/web-frameworksPR #10356. The implementation is CWIST v4.3.1, while the results dataset labels its version as4.3.There is no official single overall ranking, so the tables show requests-per-second rankings at each concurrency level.
Average latency is 4.45%, 6.22%, and 4.80% higher in GC mode, respectively. At all three concurrency levels, both implementations report zero HTTP errors, timeouts, and socket connection/read/write errors.
Data source and calculation method
d75f51932e92a8e3dc17df0c5653093314bd51c92026-11-17 06:54:33 UTCv4.3.1GET /,zrk --closed, keep-alive, approximately 15 seconds, concurrency levels 64/256/512GitHub Actions run 44035943561 for this PR provides no result artifacts. Its CWIST jobs perform only builds and 21 API contract tests, not load measurements. Accordingly, the figures below come from the complete public results dataset included in the PR, not from numbers in CI logs. The PR dataset is newer than the dataset on the base branch.
Rankings use every implementation with the same metric at the same concurrency level. Throughput is ranked from highest to lowest, while latency and standard deviation are ranked from lowest to highest. Each rank is calculated as
1 + the number of strictly better values. Some implementations have no results at concurrency levels 256 and 512, leaving denominators of 400 and 399, respectively.Values have been rounded for readability. Because these are aggregated results, total requests and received bytes may also contain fractional values or rounding. Server CPU saturation is approximately 69–71%, so these measurements should not be described as fully saturated under the project's 75% guideline.
Concurrency 64
Concurrency 256
Concurrency 512
All reactions