llama3.2:1b on A100, H100 — 2026-04-18
2026-04-18T04:32:42.000Z → 2026-04-18T04:33:12.962Z
On 2026-04-18, llama3.2:1b ran across A100, H100. H100 finished first at 590.4 tok/s, 12× faster than A100. The per-million-token cost: $1.17 (H100) vs $4.32 (A100) — the surprise: the headline GPU H100 was also the cheapest per-million tokens, coming in 3.7× less than A100.
Podium
H100
1st
thunder-h100
- peak tok/s
- 590.4
- avg tok/s
- 590.4
- $ / 1M tok
- $1.17
A100
2nd
thunder-a100
- peak tok/s
- 50.2
- avg tok/s
- 50.2
- $ / 1M tok
- $4.32