Li Benchmarks

num_cholesky greenvalidity pass

Tier 1 · wall_time · N=n=64 · variant algo_registry · OS linux

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: 0.955 (95.5% of rust speed) (table SOTA: rust — best competitor; charts use rust = 1.0)
Best competitor value: 0.0085 s

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

Problem size: N=n=64 (family num_cholesky)
Category: micro
Pillar: numerics
Package: lic
Li / catalog oracle (mean ± σ): 0.0089 ± 0.003318 s (n=86) / 0.009 ± 0.000707 s (n=101) (cpp)
Ratio vs catalog oracle: 0.9889×
Best competitor: rust (0.0085 s)
Li relative speed vs SOTA: 0.955 (1.0 = rust speed)
Validity: validity pass
Threshold: 1.2×
Source: lic/benchmarks/workloads/tier1_micro/num_cholesky

PH ids: PH-5b

Compare oracle: cpp

Relative speed vs best competitor (rust) — SOTA = 1.0, higher is better. Absolute s values are in the table below.

Language comparison
Lang	Mean ± σ	Runs	Unit	Variant	OS
li	0.0089 ± 0.003318	86	s	release	linux
cpp	0.009 ± 0.000707	101	s	release	linux
rust	0.0085 ± 0.000733	119	s	release	linux
julia	0.0096 ± 0.00119	121	s	release	linux
Metric: wall_time (value = mean of timed runs)