№ 0398GitHub
Muse Glimmer optimization reports on DGX Spark
Benchmark reports on Muse-Glimmer-30B on NVIDIA DGX Spark covering BF16 to Q4 to DFlash (a 10x speedup) and NVFP4 via SGLang, plus a head-to-head against Qwen3.6-27B.
Muse-Glimmer-30B Benchmark Reports
Benchmark and optimization reports for Muse-Glimmer-30B and comparison models on NVIDIA DGX Spark (GB10).
Reports
Glimmer Optimization Study — Full optimization journey: BF16 → Q4 kquant → DFlash speculative decoding (10× speedup), plus NVFP4 via SGLang. Includes 8-config autoresearch sweeps for both frameworks.
Glimmer vs Qwen Comparison — Head-to-head comparison with Qwen3.6-27B: decode speed, GSM8K, HumanEval pass@1, architecture impact, and use-case recommendations.
Nemotron Three-Way Comparison — Nemotron-3.5-Lightning-30B-A3B (MoE, NVFP4, vLLM) vs Glimmer and Qwen. Nemotron dominates: 95.6 tok/s (2× faster), 100% GSM8K, 92% HumanEval. Includes MoE architecture analysis and DFlash vs MTP comparison.
Key Results
| Config | Framework | Decode (tok/s) | GSM8K | HumanEval |
|---|---|---|---|---|
| Nemotron NVFP4 | vLLM | 95.6 | 100% | 92% |
| Qwen Q4+DFlash | llama.cpp | 47.3 | 92% | 92% |
| Glimmer Q4+DFlash | llama.cpp | 38.3 | 96% | 82% |
| Glimmer NVFP4 | SGLang | 11.3 | 86% | 82% |
Hardware
- NVIDIA DGX Spark (GB10, 128 GB unified memory)
- llama.cpp: CUDA 13.0, sm_121
- SGLang: FlashInfer SM120 backend, muse-glimmer branch (PR #34262)
- vLLM: 0.23.1rc1, FlashInfer 0.6.14, Marlin MoE backend
Date
August 10–12, 2026
