Skip to content

docs(worklog): the tall panel pays ~3 us of rayon region per reflector (DROPPED, recovered) - #17

Merged
shinaoka merged 1 commit into
mainfrom
docs/tall-panel-region-cost
Oct 14, 2026
Merged

shinaoka merged 1 commit into
mainfrom
docs/tall-panel-region-cost

Conversation

@shinaoka

Copy link
Copy Markdown
Member

Recovers the record from the closed PR #16. Its branch had been rebased onto main — where both of its commits became empty because the conflict in PATCHSET.md was resolved in favour of main — so the branch ended up identical to main, GitHub closed the PR, and the record and its ledger row never reached main (verified missing there).

What is pushed here is the same commit (911e897, "the tall panel pays ~3 us of rayon region per reflector"), rebased onto main (88fec8d) with the PATCHSET.md conflict resolved by keeping both rows — the merged EVD and SVD factorisation floor row and this one — which is the resolution the two changes need; neither replaces the other.

Docs only: a work-log record for the tall panel question (tlinalg-rs#46's open item), settled as DROPPED by measurement, and its ledger row.

The level-2 panel of a tall compact-WY QR is a chain of dependent reflectors, and the
fork already threads each one through the generic column split, so the level-2 work is
asked for on every reflector and answered with 1.13x (unblocked 4096x16 QR: 0.1957 /
0.1788 / 0.1731 / 0.1968 ms at 1 / 2 / 4 / 8 workers on four cores). What that costs
is a fresh rayon region per reflector: 2.11 us over the sequential product of the same
size with a hot pool, 3.83 us once a serial product separates two regions, which is
32-57 us of the panel's 173 us.

Reusing the region across the chain is a work-model change -- the ledger's Thread
policy row -- not a threshold, so this records the measurement instead of building
half of it. Also recorded: the unblocked m x 16 QR is 34% slower threaded at m=1024
and 17% at 2048, so the rank-one arm's 8192-element floor is active through a losing
range; that floor is load-bearing where it was measured and was not moved.

No code change. Record: docs/worklogs/2026-10-14-tall-panel-region-cost.md
@shinaoka
shinaoka merged commit bfe4041 into main Oct 14, 2026
7 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant