[CUDA][Reduce] Add per-reduce control for FP32x2 accumulation (#3057)
Add per-reduce CUDA FADD2 control
L
Lei Wang committed
ddf3bdcb7b3b8eb6bb5afeeec23b1dfea53ebf95
Parent: d72e1d4
Committed by GitHub <noreply@github.com>
on 8/21/2026, 5:18:06 AM