Can you run Qwen Next on a 3090 + 64GB CMP 170HX? Local deployment help
JustinPooDough · reddit · 2026-09-07
A user with a 3090 (24GB) plus an unlocked CMP 170HX (64GB) and 64GB DDR4 asks whether any configuration can run Qwen 3.8 Next at acceptable speeds, factoring in the CMP 170HX's PCIe 2.0 x4 bottleneck. They've already run Qwen 3.7 27B and MiniMax H3 at BF16 on it, but the larger Qwen Next may require DRAM offloading.
More from Infra
- Lightpanda, a Zig-based headless browser built for AI, reaches 34.6k GitHub stars — lightpanda-io · 2026-09-07
- CPU Cache video series shared, with Scott Meyers' timeless memory hierarchy talk — blaizedsouza · 2026-09-07
- GPU engineer: stop saying CUDA Core/Tensor Core, they're marketing terms obscuring SM microarchitecture — yacineMTB · 2026-09-07
- Updated guide: running GitHub Copilot with local models via VS Code and Lemonade — admcpr · 2026-09-07
- Tech companies eye Argentina's windswept Patagonia for massive data centers — Shot-Height-7194 · 2026-09-07
- Microsoft open-sources tgrep, a trigram-indexed grep up to 52x faster than ripgrep — jedisct1 · 2026-09-07