From 93e24f29f4984ad98a8cedb2a2ffa633a5a48ec6 Mon Sep 17 00:00:00 2001 From: laurent Date: Mon, 26 Jun 2023 22:01:29 +0100 Subject: [PATCH] Add the f16 sum kernel. --- kernels/src/reduce.cu | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/kernels/src/reduce.cu b/kernels/src/reduce.cu index d12d6b22..76dbbdc3 100644 --- a/kernels/src/reduce.cu +++ b/kernels/src/reduce.cu @@ -43,6 +43,10 @@ extern "C" __global__ void FN_NAME( \ } \ } \ +#if __CUDA_ARCH__ >= 530 +SUM_OP(float, sum_f32) +#endif + SUM_OP(float, sum_f32) SUM_OP(double, sum_f64) SUM_OP(uint32_t, sum_u32)