Qwen 3.8 27B quantization achieves 200k context on 16GB VRAM
abskvrm · reddit · 2026-08-28
A user achieved over 200k context length with the Qwen 3.8 27B UD-IQ3XXS model on 16GB VRAM. Compared to the previous UD-Q3KXL setup, prompt processing speed dropped from 700-800 tk/s to 400 tk/s, while quality differences are yet to be fully tested. KV cache was quantized to q51.
Related event: Qwen 3.8 27B Quantization Tests: 200K Context on 16GB VRAM(4 posts)→
More from Infra
- Selling B300 inference nets $40/hr vs $6/hr rental — but the math is disputed — TheZachMueller · 2026-10-15
- ASML CFO says US Match Act restricting DUV sales to China would backfire — pstAsiatech · 2026-10-15
- ASML sold out through 2028 but stock dips 2%; TL20 index up 57% YTD — TiernanRayTech · 2026-10-15
- Bloomberg breakdown: a 1GW Texas AI data center may need nearly $3B for overruns alone — rohanpaul_ai · 2026-10-15
- 7 System Design Algorithms to Know: Merkle Trees to CAP/PACELC — goyalshaliniuk · 2026-10-15
- MiniCPM-V 4.7 runs a 35B MoE multimodal model on phones with just 3GB memory — nevrekaraishwa2 · 2026-10-15