NInfer vs llama.cpp vs vLLM: rigorous quality and speed test of Qwen3.8-27B NVFP4 on one RTX 5090
bengizmoed · reddit · 2026-09-05
A developer shares a rigorously controlled comparison of three inference engines serving Qwen3.8-27B NVFP4 on a single RTX 5090 (32GB, eGPU), evaluated on 6 tiers of real production workloads — relevance classification, needle retrieval at 64K-240K context, multi-transcript QA, reasoning, structured extraction, and agent tool replay — with paired prompts, seeds, gold labels, and cluster-bootstrap CIs. A 4-model review panel (Codex/GPT-5, DeepSeek V4 Pro, Grok 4.5, Kimi K3) audited the methodology and caught 10 issues including 4 mislabeled gold items.
Key findings:
- Quality is statistically indistinguishable across all three engines; needle retrieval is 100% on every item that fits
- NInfer's retrieval edge comes entirely from context capacity (240K vs llama.cpp's 196K), not quality
- All engines fail tool replay equally; llama.cpp is limited to parallel=1
- NInfer offers MTP3 (76% acceptance) with x2 concurrency; VRAM use is 29.6-31.6GB across engines
Takeaway: pick based on concurrency and context needs, not quality fears.
More from Infra
- llama.cpp team launches Llama: one-click local AI on your Mac menu bar — Hugging Face · 2026-10-23
- China's daily token use hits 500T, up 4x as average price doubles to ¥13.46 — teortaxesTex · 2026-10-23
- New all-reduce kernels deliver 12% decode speedup for DeepSeek-V4.1 on PCIe-connected GPUs — AlpinDale · 2026-10-23
- Offshore oil rigs could be converted into AI data centers powered by stranded gas — clemnt · 2026-10-23
- Same native UI, two stacks: Swift build is ~80MB Mac-only, Zig build is ~10MB cross-platform — jedisct1 · 2026-10-23
- SurfSense: an open-source local NotebookLM with MCP agents and a prompt-caching lesson — FurtiveMirth · 2026-10-23