-------------------------------------------------------------------------------- Running performance on file test\p30_deffer_impl_flat.ps -------------------- NV40 -------------------- Target: GeForce 6800 Ultra (NV40) :: Unified Compiler: v65.04 IPU0 ------ Simplified schedule: -------- Pass | Unit | uOp | PC: Op -----+--------+------+------------------------- 1 | SCT0 | div | 0: TEXh h1, f[TEX1], TEX0; | SCT1 | mov | 1: NRMh h4.xyz, f[TEX3]; | TEX | tex | 0: TEXh h1, f[TEX1], TEX0; | SRB | nrm | 1: NRMh h4.xyz, f[TEX3]; | SCB1 | mul | 2: MOVh h0.w, const; | | | 2 | SCT1 | mul | 4: MOVh h4.w, h1; | SCB0/1 | mad | 5: MADh h6, h1.xyzx, const.xxxy, const.yyyz; | | | 3 | SCT0 | div | 7: MADh h0.xyz, h4, const.xxx-, f[TEX2]; | SCB0 | mad | 7: MADh h0.xyz, h4, const.xxx-, f[TEX2]; Pass SCT TEX SCB 1: 50% 100% 25% 2: 25% 0% 100% 3: 75% 0% 75% 4: 0% 0% 0% MEAN: 37% 25% 50% Pass SCT0 SCT1 TEX SCB0 SCB1 1: 100% 0% 100% 0% 100% 2: 0% 100% 0% 100% 100% 3: 100% 0% 0% 100% 0% 4: 0% 0% 0% 0% 0% MEAN: 50% 25% 25% 50% 50% Cycles: 4.00 :: R Regs Used: 3 :: R Regs Max Index (0 based): 3 Max register used is > number of registers used, registers are not being used efficiently -------------------------------------------------------------------------------- Running performance on file test\p30_deffer_impl_flat.ps -------------------- NV40 -------------------- Target: GeForce 6800 Ultra (NV40) :: Unified Compiler: v81.95 Cycles: 2.00 :: R Regs Used: 3 :: R Regs Max Index (0 based): 3 Max register used is > number of registers used, registers are not being used efficiently Pixel throughput (assuming 1 cycle texture lookup) 3.20 GP/s -------------------------------------------------------------------------------- Running performance on file test\p30_deffer_impl_flat.ps -------------------- G70 -------------------- Target: GeForce 7800 GT (G70) :: Unified Compiler: v81.95 Cycles: 2.00 :: R Regs Used: 3 :: R Regs Max Index (0 based): 3 Max register used is > number of registers used, registers are not being used efficiently Pixel throughput (assuming 1 cycle texture lookup) 4.80 GP/s