+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Optimization & Performance (https://www.datasculpting.ai/inference/forumdisplay.php?fid=10)
+--- Forum: Performance Tuning (https://www.datasculpting.ai/inference/forumdisplay.php?fid=42)
+--- Thread: what changed after the inference shift with MLX (/showthread.php?tid=1483)
Pages:
1
2
Re: what changed after the inference shift with MLX - deltagrid - 10-18-2026
Ran a quick check on my box. I couldn't reproduce the slowdown until I pushed context much higher. The surprise was that memory pressure showed up before throughput collapsed.
Re: what changed after the inference shift with MLX - stackarch - 10-18-2026
Appreciate the detail. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
Re: what changed after the inference shift with MLX - kernelrunner - 10-18-2026
Quote:Appreciate the detail. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one more optimi
same here on my side.
Re: what changed after the inference shift with MLX - rocmpath - 10-18-2026
Quote:Ran a quick check on my box. I couldn't reproduce the slowdown until I pushed context much higher. The surprise was that memory pressure showed up bef
Ran a quick check on my box. The issue went away once I changed one setting and restarted the stack cleanly. The surprise was that memory pressure showed up before throughput collapsed.
Re: what changed after the inference shift with MLX - rocmhouse90 - 10-18-2026
Quote:Ran a quick check on my box. On 4090 with PyTorch, I saw basically the same behavior. The surprise was that memory pressure showed up before throughpu
Tried to reproduce it just now. The issue went away once I changed one setting and restarted the stack cleanly. The surprise was that memory pressure showed up before throughput collapsed.
Re: what changed after the inference shift with MLX - papermixer4 - 10-18-2026
I'd approach it in a boring order. On dual 3090, the easiest win is often to back off the aggressive settings and see whether the problem survives.
Re: what changed after the inference shift with MLX - kernelmike - 10-18-2026
Quote:[quote]Ran a quick check on my box. I couldn't reproduce the slowdown until I pushed context much higher. The surprise was that memory pressure showed
Not sure I agree there. I get why that seems right, but on B200 the bottleneck felt elsewhere. Could be a workload difference, but I wouldn't generalize too fast from one setup.
Re: what changed after the inference shift with MLX - sanepeak48 - 10-18-2026
Did a small comparison this afternoon. On A6000 with vLLM, I saw basically the same behavior. The surprise was that memory pressure showed up before throughput collapsed. At this point the real bottleneck is my patience.
Re: what changed after the inference shift with MLX - cudapeak - 10-19-2026
Quote:Did a small comparison this afternoon. On A6000 with vLLM, I saw basically the same behavior. The surprise was that memory pressure showed up before t
A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Mistral and compare from there.
Re: what changed after the inference shift with MLX - macropilot - 10-19-2026
I hit almost the same issue last night.