Li Benchmarks

matmul_blocked redvalidity pass

Tier 1 · wall_time · N=512 block=64 · OS linux

Validity gate

passpass — perf claims allowed when ratio is green

Source: latest.csv:perf_present

Wall-clock row exists only — no explicit correctness signal yet.

Ratio vs best competitor: 0.639 (63.9% of julia speed) (best in series: julia — Li is never labeled best)
Best competitor value: 0.0101 s

Numeric oracle

No analytical verify rows in ingest yet. Run tier-1 with --verify so latest.csv exports verify_ulps / verify_within_1ulp.

Problem size: N=512 block=64
Category: micro
Pillar: numerics
Package: lic
Li / catalog oracle (mean ± σ): 0.0158 s / 0.0102 s (cpp)
Ratio vs catalog oracle: 1.5490×
Best competitor: julia (0.0101 s)
Li relative speed vs SOTA: 0.639 (1.0 = julia speed)
Validity: validity pass
Threshold: 1.2×
Source: lic/benchmarks/tier1_micro/matmul_blocked

PH ids: PH-5b

Performance vs best competitor

Compare oracle: cpp

Relative speed vs best competitor (julia) — SOTA = 1.0, higher is better. Absolute s values are in the table below.Validity gate not passed — bars are informational only.

julia (release)SOTA
1.000
cpp (release)
0.990
rust (release)
0.981
li (release)
0.639

Absolute measurements

Language comparison
Lang	Mean ± σ	Runs	Unit	Variant	OS
li	0.0158	—	s	release	linux
cpp	0.0102	—	s	release	linux
rust	0.0103	—	s	release	linux
julia	0.0101	—	s	release	linux
Metric: wall_time (value = mean of timed runs)

Host OS

Measurement hosts by language
Scope	OS
Row aggregate	linux
li (release)	linux
cpp (release)	linux
rust (release)	linux
julia (release)	linux

Latest history deltas

ratio_vs_cpp: 1.5347 → 1.549 (Δ 0.0143) · regressed

All latest deltas →

← Overview