-
Notifications
You must be signed in to change notification settings - Fork 0
Performance Comparison
Simon B.Stirling edited this page Mar 2, 2026
·
4 revisions
I generated this snapshot automatically with tests/benchmark_compare.sh.
- generated_utc:
2026-03-02T14:02:12Z - host:
runnervmnay03 - kernel:
Linux 6.14.0-1017-azure x86_64 - l0c:
./bin/l0c - gcc:
gcc (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0
I compare process-level throughput (ops/s) for equivalent minimal tasks:
- build throughput: compile/build a minimal add kernel/program repeatedly
- run throughput: execute add with two decimal args repeatedly
This is an operational comparison, not a language-runtime microbenchmark.
| Workload | L0 (l0c) ops/s |
GCC C ops/s |
|---|---|---|
| Build minimal add artifact | 1242 | 20 |
| Run minimal add artifact/program | 1560 | 1076 |
- I run this on the local host, so values are machine-dependent.
- CI smoke/full perf gates remain the enforcement source for regressions.
- How-To-Write-L0
- Language-Reference
- Instruction-Set
- CLI-and-Compiler-Spec
- Implementable-Spec
- Command-Reference
- Examples-Catalog
- LLM-Quick-Reference
- Opcode-Examples
- LLM-Doc-Index