Wow! Qwen 3.6:35b-a3b on a 3090... pretty amazing.
Summary
A user shares impressive results running a quantized Qwen 3.6:35b-a3b model on a used RTX 3090, achieving 160 tokens per second output after fitting the model into VRAM, and demonstrates vision capabilities with a 75-second video processing time.
Similar Articles
Ternary Qwen3.6 27B Tested on 3090!
User tests ternary quantized Qwen3.6 27B on an RTX 3090, achieving 60 tk/s with two slots and 100k KV cache using 21GB VRAM, with good quality and stable tool calls.
I pushed Qwen3.8-27B to 99 tps single request and 1150 tps with a batch request on a RTX 3090
The author optimized the Qwen3.8-27B model inference on an RTX 3090 GPU, achieving up to 99 tokens per second for single requests and 1150 tps with batch processing through various quantization and optimization techniques, and released the updated code on GitHub.
@rohanpaul_ai: Beautiful visual of somebody running, qwen 3.8 27B locally on a RTX 5090 32 GB VRAM system with 115 tokens/sec note, Qw…
Tweet highlights running the Qwen 3.8 27B model locally on an RTX 5090 system with 32GB VRAM, achieving 115 tokens/sec, and notes the official BF16 checkpoint is 55.6GB.
Qwen 3.6 27B Speculative Decoding Bench: Pushing ~100 TPS on a single RTX 3090
A detailed benchmark comparing speculative decoding engines for Qwen 3.6 27B on a single RTX 3090, showing ik_llama achieving ~100 tokens per second in code generation. Results include decode TPS, TTFT, VRAM usage, and context degradation across 5 engine variants.
@ItsmeAjayKV: Update on 3090: Now with Qwen 3.6-35b-a3b moe (q6_k_xl). Crossed 90 t/s for the very first time, no MTP yet, prefill sp…
A user reports achieving over 90 tokens per second inference speed with Qwen 3.6-35b-a3b MoE model on an RTX 3090 using llama.cpp, with prefill speeds exceeding 1000 t/s, indicating practical local deployment of large language models on consumer hardware.