Skip to content

[None][perf] GVR top-K decode: enable R0 histogram-ladder admission by default #16457

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

Open
longcheng-nv wants to merge 20 commits into
base: main
Choose a base branch
from
+1,974 −47
Open

[None][perf] GVR top-K decode: enable R0 histogram-ladder admission by default #16457

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
20 commits
Select commit Hold shift + click to select a range
ad3f9c4
[None][perf] gvr top-k decode: add block_count_ge_multi + R0 knobs (i…
longcheng-nv Jul 15, 2026
9878e1a
[None][perf] gvr top-k decode: add P1b histogram-quantile rung placem…
longcheng-nv Jul 15, 2026
d2252ee
[None][perf] gvr top-k decode: R0 single-pass admission in _run_phase…
longcheng-nv Jul 15, 2026
af5bd95
[None][perf] gvr top-k decode: inline log-falsi R1 shot on R0 miss
longcheng-nv Jul 15, 2026
b28b884
[None][perf] gvr top-k decode: fb_fix bounded log-falsi fallback (rep…
longcheng-nv Jul 15, 2026
a8c69a7
[None][perf] gvr top-k decode: cluster-aware R0 (cs>1 long-row branch)
longcheng-nv Jul 15, 2026
4d00c8a
[None][perf] gvr top-k decode: R0 dispatch defaults ((0.85,0.35) qfra…
longcheng-nv Jul 15, 2026
f926cdf
[None][perf] gvr top-k decode: p1b_cache path (SMEM gather-cache, 16-…
longcheng-nv Jul 15, 2026
a489d80
[None][perf] gvr top-k decode: fix cluster-R0 admission/refine handof…
longcheng-nv Jul 15, 2026
1ba38ec
[None][perf] gvr top-k decode: seeded cluster R0-miss refine (fold R1…
longcheng-nv Jul 15, 2026
c528253
[None][perf] gvr top-k decode: add fused rank-and-scatter P4 (inert f…
longcheng-nv Jul 15, 2026
6002551
[None][perf] gvr top-k decode: enable rank-scatter P4 by default unde…
longcheng-nv Jul 15, 2026
39f17b2
[None][perf] gvr top-k decode: p1b_cache ON for all dtypes at cluster…
longcheng-nv Jul 15, 2026
3e20622
[None][chore] GVR top-K decode: ruff-format the R0 additions
longcheng-nv Jul 15, 2026
95a8388
[None][perf] GVR top-K decode: enable R0 admission by default + equiv…
longcheng-nv Jul 15, 2026
0182519
[None][perf] GVR top-K decode: pick_config launch-shape policy + laun…
longcheng-nv Jul 16, 2026
88a563b
[None][perf] GVR top-K decode: virtual seed rung (r0_vseed) + per-K r…
longcheng-nv Jul 16, 2026
eae3745
[None][fix] GVR top-K decode: exact tie resolution in the P4 straddli…
longcheng-nv Jul 16, 2026
af7351c
[None][chore] Fix ruff-format pre-commit violations
longcheng-nv Jul 17, 2026
7408f3b
[None][test] Preserve -1 encoding for true index 0 in R0 pre_idx helper
longcheng-nv Jul 17, 2026
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
Loading
Loading