Local tuning of GLM-5.3-Flash on 2×GB10 lifts decode from 49 to 58 tok/s
EAccelerate_42 · x · 2026-10-16
The author shares September tuning results running GLM-5.3-Flash locally on 2× GB10 (GX10 + DGX Spark): 32K prefill improved from 1,426 to 1,700 tok/s, and coding decode from 49 to 58 tok/s (18% gain). They note these are historical results from their own setup, not a universal benchmark, and ask what bottlenecks others are chasing.
More from Infra
- Google 将 2026 年资本开支上调至最高 2050 亿美元 — emmanuelvivier · 2026-10-16
- Google open-sources AX, a Kubernetes-style orchestrator for billions of AI agent tasks — thisdudelikesAI · 2026-10-16
- Valtz v0.1.0.11 adds FLUX.3 support, local image gen up to 4K on Mac — TgoAI · 2026-10-16
- Baidu open-sources FlashMask: up to 1.55x faster than FlashAttention-4 for long-context training — BaiduResearch · 2026-10-16
- GYRO quant of Qwen3.8-Flash-Next tested: near Q3 quality at much smaller size — PlasticRevenue4601 · 2026-10-16
- RL train-sample systems are frontier AI's hardest infra problem — and a trade secret — brianryhuang · 2026-10-16