Go is simply not feasible for CUDA work mainly because of the go runtime that manages GC, allocation, scheduling etc

GPU kernels want explicit memory control and as little go-runtime like overhead as possible