+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Drivers / Frameworks / Toolkits (https://www.datasculpting.ai/inference/forumdisplay.php?fid=8)
+--- Forum: AMD / ROCm (https://www.datasculpting.ai/inference/forumdisplay.php?fid=32)
+--- Thread: HIP kernel compile issue for Mixtral (/showthread.php?tid=1336)
Pages:
1
2
Re: HIP kernel compile issue for Mixtral - rocmriley - 10-19-2026
I don't think that's the whole story. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Worth testing the simpler explanation before adopting the more interesting one.
Re: HIP kernel compile issue for Mixtral - rocmriley - 10-19-2026
I don't think that's the whole story. I've seen people blame prefix caching when the real issue was much more mundane. Could be a workload difference, but I wouldn't generalize too fast from one setup.
Re: HIP kernel compile issue for Mixtral - phasetrack - 10-19-2026
Maybe, but I had the opposite result. I get why that seems right, but on 7900 XTX the bottleneck felt elsewhere. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.
Re: HIP kernel compile issue for Mixtral - rocmriley - 10-19-2026
My first guess would be this: Before rebuilding anything, I'd test one clean baseline with Mixtral and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.