Can a 3090+3060 rig with 64GB RAM run Flash Next and the upcoming Qwen4?
No-Yak4416 · reddit · 2026-10-22
Running Qwen3.8-27B on dual GPUs (3090 + 12GB 3060) with 32GB RAM at 40 TPS via MTP, the poster weighs upgrading to 64GB RAM to run Flash Next and the upcoming Qwen4.
The open question: after desktop OS overhead and full context, can that RAM+VRAM split actually fit those models, and is the upgrade worth it? Comments cover the practical math of local deployment configs.
More from Infra
- SpaceX's Bastrop site balloons into industrial city with 1.1M sq ft solar factory — XFreeze · 2026-10-22
- How to benchmark compiler opts: inspect emitted code, split cold vs steady — cfregly · 2026-10-22
- Nested CUDA graphs cut replay launches from 16 to 2 per call — cfregly · 2026-10-22
- Nested graphs deliver 4.5× steady-state speedup in depth-8 benchmark — cfregly · 2026-10-22
- Reddit asks: which 50-150B MoE models are worth running locally besides Qwen Flash Next — FlyingFishMakeAWish · 2026-10-22
- Open-source tool condenses millions of logs into an LLM-sized snapshot — CommanderAlPachino · 2026-10-22