Hi,
PIX tool on RDNA3 (20 Work Group Processors) shows 1836 CSBusyCycles for below compute shader. could someone explain me how the cycles has been computed?
struct MeshletData
{
float4 a;
};
RWStructuredBuffer<MeshletData> BufferOut : register(u0);
[numthreads(32, 1, 1)]
void CSMain( uint3 id : SV_DispatchThreadID )
{
float4 t1 = vector<float, 4>(id.x,id.y,id.z,id.x+id.z);
float4 t2 = vector<float, 4>(1.232, 1.456, 1.789, 1.91);
for(uint i=0u; i<32u; i++) {
t1 = (t1*t2);
}
BufferOut[id.x].a = t1;
}
Thanks,
Venkatesh.
