+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Drivers / Frameworks / Toolkits (https://www.datasculpting.ai/inference/forumdisplay.php?fid=8)
+--- Forum: NVIDIA / CUDA (https://www.datasculpting.ai/inference/forumdisplay.php?fid=31)
+--- Thread: Tensor cores actually helping here on B200 (/showthread.php?tid=673)
Pages:
1
2
Re: Tensor cores actually helping here on B200 - dockember - 10-06-2026
My first guess would be this: On dual 3090, the easiest win is often to back off the aggressive settings and see whether the problem survives.
Re: Tensor cores actually helping here on B200 - amberpatch25 - 10-06-2026
I think the bottleneck is somewhere else. I get why that seems right, but on mini PC the bottleneck felt elsewhere. Worth testing the simpler explanation before adopting the more interesting one.
Re: Tensor cores actually helping here on B200 - kernelmike - 10-07-2026
Can you fill in a bit more context? What model, what runtime, and roughly what context size?
Re: Tensor cores actually helping here on B200 - orbitbench6 - 10-10-2026
I'd start with the simpler stack.