vortex-data/vortex - CodSpeed

vortex

Blog Docs Changelog

Performance History

Latest Results

Measure chunked take with walltime Signed-off-by: Connor Tsui <connor.tsui20@gmail.com>

ct/codspeed-repeatability

1 minute ago

Support decimal values in Widen Keep the logical decimal dtype while storing values in a narrower decimal child. Execute into the logical dtype's canonical buffer width and serialize narrow leaves through the legacy decimal wire format without executing them. Signed-off-by: Connor Tsui <connor.tsui20@gmail.com>

ct/decimal-widen

23 minutes ago

Warm allocator metadata before chunked take simulation Signed-off-by: Connor Tsui <connor.tsui20@gmail.com>

ct/codspeed-repeatability

23 minutes ago

Restore practical benchmark examples and measurement guidance Signed-off-by: Connor Tsui <connor.tsui20@gmail.com>

ct/codspeed-repeatability

39 minutes ago

Keep benchmark options fixed in rebuild comparisons Signed-off-by: Connor Tsui <connor.tsui20@gmail.com>

ct/codspeed-repeatability

47 minutes ago

Stabilize CodSpeed microbenchmark reporting Signed-off-by: Connor Tsui <connor.tsui20@gmail.com>

ct/codspeed-repeatability

1 hour ago

Mark benchmark inputs as optimized before timing Execution now optimizes each array that it reaches unless the array is already marked. Arrays that a reader decodes are marked, but benchmarks build their inputs with constructors. Thus the timed region also timed the optimizer, and CodSpeed reported about 400 regressions that production does not see. test_harness::mark_optimized optimizes an input during setup and panics if the optimizer rewrites it, so a benchmark cannot time a different array than the one it built. test_harness::mark_decoded marks each node without trying any rule, as decoding does. The cosine similarity benchmark uses it for an extension array over constant storage, which a file can hold but the optimizer folds into a constant. Lazy arrays that a benchmark builds inside the timed closure stay unmarked, because their optimization is part of the cost it measures. Signed-off-by: Robert Kruszewski <github@robertk.io>

rk/optimize-execute-loop

1 hour ago

refactor(jni): batch independent buffer ownership metadata Signed-off-by: Christopher Wirt <cwirt@quantcast.com>

quantcast:cwirt/upstream-jni-independent-buffers

2 hours ago

Latest Branches

-1%

Support decimal values in Widen#10946

29 minutes ago

d2eb3da

ct/decimal-widen

+1%

Stabilize CodSpeed microbenchmarks#11110

26 minutes ago

6dc7070

ct/codspeed-repeatability

-16%

Optimize arrays bottom up and as execution reaches them#11090

4 hours ago

4db3b2a

rk/optimize-execute-loop

© 2026 CodSpeed Technology

Home Terms Privacy Docs