Qwen3.8 Optimization: Disk Streaming Yields 40 tok/s
samfajreldines · x · 2026-08-29
EyalToledano shared a breakthrough in optimizing Qwen3.8-Flash-Next. By storing 60% of experts on disk and streaming them to memory on-demand (similar to n-gram streaming), the full expert model runs on just 37GB of memory on an M4 Max, decoding at 40 tok/s.
Related event: Optimization Lets Qwen3.8-Flash-Next Run on 37GB VRAM(4 posts)→
More from Infra
- Perplexity's new embedding models: index with a 9B model, search with 0.6B — antoine_chaffin · 2026-10-15
- Unsloth ships GGUF quants for Nemotron 3.5: Bolt-74B runs in 24GB RAM — danielhanchen · 2026-10-15
- vLLM-Omni ships Day-0 serving for FLUX 3 Image Klein with BF16, FP8, NVFP4 support — vllm_project · 2026-10-15
- NVIDIA's Nemotron 3.5 Super VL lands training and inference support on Applied Compute — ypatil125 · 2026-10-15
- CoreWeave bond yield tops 16% as loans outlast customer contracts by ~2 years — tengyanAI · 2026-10-15
- AGI isn't real until H100 hourly rates jump 10x, argues compute economics essay — corbtt · 2026-10-15