cornerstonejs/codecs - CodSpeed

codecs

Blog Docs Changelog

Performance History

No performance history available yetOnce you have some commits, you will be able to see the performance history of your primary branch.

Latest Results

Update to newer test benchmark

fix/htj2k-partial

7 hours ago

Update to newer test benchmark

fix/htj2k-partial

7 hours ago

ci: retrigger codspeed-bench after main baseline seeded #67 landed on main and produced the first CodSpeed baseline (run 26167789689). Re-running codspeed-bench on this PR so the sticky comment can flip from "Congrats!" to a real before/after table with the fake-regression deltas highlighted on every codec. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

regression-fake

17 hours ago

ci: retrigger codspeed-bench after main baseline seeded #67 landed on main and produced the first CodSpeed baseline (run 26167789689). Re-running codspeed-bench on this PR so the sticky comment can flip from "Congrats!" to a real before/after table with the fake-regression deltas highlighted on every codec. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

regression-fake

17 hours ago

ci: seed CodSpeed baseline on main pushes and workflow_dispatch (#67) Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

main

18 hours ago

feat: add artificial CPU burn for benchmark regression detection Introduced a block of code across multiple decoder and encoder implementations to create a controlled CPU load, allowing for the detection of performance regressions in the benchmark pipeline. This block is marked for removal before merging.

regression-fake

18 hours ago

feat: add artificial CPU burn for benchmark regression detection Introduced a block of code across multiple decoder and encoder implementations to create a controlled CPU load, allowing for the detection of performance regressions in the benchmark pipeline. This block is marked for removal before merging.

regression-fake

18 hours ago

docs(bench): explain the CPU-simulation measurement model Adds BENCHMARKING.md at the repo root covering: - why we use mode: simulation (Cachegrind) instead of walltime - how the modeled instruction-time numbers relate to real wall-clock (JS loops inflate 30-100x, wasm decode 5-15x) - how to read each of the four bench types we emit per codec (instantiate+destroy, cold, warm, dispatcher integration) - what the dashboard warnings mean and which to ignore (syscall-attribution blowup, anonymous V8 JIT frames, V8 tier-up artifacts on instantiation benches) - run-to-run variance characteristics observed across three identical-source runs - how to add a new bench Cross-references the doc from: - README.md (new Benchmarking section, before Codec Package Anatomy) with a TL;DR pointing at BENCHMARKING.md - .github/workflows/pr-checks.yml (expanded comment above the `mode: simulation` line explaining why we chose it and what the trade-offs are) No behaviour change; pure documentation of what we built up in commits a5b8ee6..b4b14d1.

add-tests

1 day ago

Latest Branches

-4%

fix: Decode partial htj2k stream#68

7 hours ago

ff08a16

fix/htj2k-partial

-64%

test(bench): inject fake CPU regression to validate detection pipeline#66

17 hours ago

2c1ff1c

regression-fake

N/A

test: migrate to vitest with parallel per-package CI + benchmark PR comments#64

1 day ago

7fee128

add-tests

© 2026 CodSpeed Technology

Home Terms Privacy Docs