Same model, different hardware: dev finds A100s vs B300s produce different outputs
Kyrannio · x · 2026-10-30
Developer Kyrannio may publish a comparison showing how hardware affects his posttrained model's outputs, noting A100s vs B300s produce noticeably different results.
He also found that render-time optimizations like caching reference inputs often degrade quality — and that even without speed/quality tradeoffs, hardware type itself measurably changes outputs.
More from Infra
- Anthropic eyes $2T valuation on $100B raise, with $431B compute commitments and circular vendor financing — LexSokolin · 2026-10-30
- Cloudflare private origins goes GA: WAF and caching for internal APIs and MCP servers — Cloudflare Blog · 2026-10-30
- Pika H3 Flex ported to Mac via MLX: 5s video in ~5 minutes on M4 Max — Opening-Ad5541 · 2026-10-30
- FlashAttention-2 v2.9.2 Windows wheels released for PyTorch 2.15.0+cu134 — Zestyclose_Bake3680 · 2026-10-30
- AgentX benchmark: lossless-CSF hits 11.4M KV cache tokens, 8.7x DeepSeek V4.1 Flash baseline — TheZachMueller · 2026-10-30
- 'Girls only want one thing': an $496,639 8x NVIDIA B300 deep learning server — Kyrannio · 2026-10-30