Tensor cores actually helping here with SGLang

[rasterpoint]

I'd approach it in a boring order. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at speculative decoding. Once you know the plain path works, then add the faster path back in piece by piece.

[bytelab]

Not sure I agree there. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Could be a workload difference, but I wouldn't generalize too fast from one setup.