Visitar URL original
Anthropic’s original performance take-home, now a community leaderboard. Hand-optimize a kernel for a simulated wide-issue VLIW machine and chase the lowest cycle count.
Your perf_takehome.py builds the instruction stream; the judge scores the worst cycle count across nine seeded validation runs — lower is better. Model reference runs on the board show where the frontier sits. Check out the GitHub repo to get started.
Community project — not affiliated with or endorsed by Anthropic.
Ranked by clock cycles (lower is better) — each account’s best score counts
| Rank | Author | Cycles | Attempts |
|---|---|---|---|
| #1 | @YuleHou | 864 | 10 |
| #2 | @josusanmartin | 864 | 359 |
| #3 | @ryan_kirkman | 865 | 14 |
| #4 | @ChangyeLi03 | 868 | 1 |
| #5 | @HaydenCc51623 | 869 | 13 |
| #6 | @SaifAlHarthi | 870 | 63 |
| #7 | @LigengZhu | 875 | 6 |
| #8 | @danielpiers | 875 | 1 |
| #9 | @jurajselep | 876 | 1 |
| #10 | @PfistererF66133 | 878 | 6 |
Model rows are reference runs shown for orientation only — single data points under specific harness conditions. Replication attempts may yield different results.