Oleksandr Pavlyk 2a515c2569 Change in how FAST/SLOW deciision is arrive at ...

Now:

  - establish a candidate clear timing gap from summary timing intervals, as before
  - if bulk sample times and frequencies are available on both sides,
    compute cycles = time * frequency
  - derive bulk cycle intervals from min/q1/median/q3
  - confirm the gap direction from those bulk cycle intervals
  - only fall back to summary sm_clock_rate_mean confirmation when bulk cycle data
    is unavailable

  I also split the reason codes so the evidence source is visible:

  - clear_gap_confirmed_by_bulk_cycles
  - bulk_cycle_gap_not_confirmed
  - clear_gap_confirmed_by_summary_cycles
  - summary_cycle_gap_not_confirmed

Updated tests in python/test/test_nvbench_compare.py cover both the bulk-confirmed
and bulk-rejected paths, along with the renamed summary reason codes.

2026-06-03 15:57:34 -05:00

..

nvbench_json

Build multi architecture cuda wheels (#302)

2026-01-29 01:13:24 +00:00

__init__.py

Build multi architecture cuda wheels (#302)

2026-01-29 01:13:24 +00:00

nvbench_compare.py

Change in how FAST/SLOW deciision is arrive at

2026-06-03 15:57:34 -05:00

nvbench_histogram.py

Build multi architecture cuda wheels (#302)

2026-01-29 01:13:24 +00:00

nvbench_json_summary.py

Provide BenchmarkResult class for parsing JSON output of NVBench-instrumented benchmarks (#356)

2026-05-13 13:23:58 -05:00

nvbench_plot_bwutil.py

Apply reviewer feedback

2026-02-26 19:23:51 +01:00

nvbench_walltime.py

Build multi architecture cuda wheels (#302)

2026-01-29 01:13:24 +00:00

test_cmp.json

Build multi architecture cuda wheels (#302)

2026-01-29 01:13:24 +00:00

test_ref.json

Build multi architecture cuda wheels (#302)

2026-01-29 01:13:24 +00:00