## ncu_5120x10240.ncu-rep
served_M8    void exl3_gemm_kernel<4, 0, 2, 16, 32, 128, 4, 3> dur=  36.96us dram= 72.7% (  713GB/s rd=  26.4MB) sm= 53.9% issue= 53.9% tensor= 15.0% alu= 51.4% fma= 20.4% occ= 33.7/  33% l2hit= 23.3% regs=107 smem= 91.0K grid=128x512 bankcf=20420 stalls=barrier:1.7,wait:1.3,math_pipe_throttle:1.2,selected:1.0
mma_M8       void ag_mma_kernel<4, 2, 1, 1, 1>(const void *, const unsign dur=  38.24us dram= 70.3% (  690GB/s rd=  26.4MB) sm= 39.8% issue= 39.8% tensor=  6.9% alu= 38.7% fma= 16.1% occ= 55.6/  67% l2hit= 33.0% regs=64 smem=  9.0K grid=800x256 bankcf=435200 stalls=long_scoreboard:8.8,wait:1.6,not_selected:1.1,selected:1.0
up_gemv_M8   void exl3_gemv_kernel<4, 0, 2, 1, 1, 0> dur=  35.58us dram= 75.4% (  741GB/s rd=  26.4MB) sm= 26.9% issue= 26.9% tensor= 14.3% alu= 21.6% fma= 14.4% occ= 21.2/  50% l2hit= 31.4% regs=80 smem= 17.0K grid=160x256 bankcf=61440 stalls=long_scoreboard:4.7,barrier:1.1,selected:1.0,wait:0.8
simt_M8      void ag_gemv_kernel<Exl3GemvLoader<4, 2, 8>, 4, 0, 8, 4>(con dur=  55.36us dram= 48.5% (  476GB/s rd=  26.4MB) sm= 86.6% issue= 60.6% tensor=  0.0% alu= 20.5% fma= 36.9% occ= 50.0/  67% l2hit= 24.2% regs=64 smem= 21.0K grid=640x256 bankcf=7508 stalls=mio_throttle:2.7,not_selected:2.1,selected:1.0,barrier:0.9
simt_M8      void ag_gemv_reduce_kernel<Exl3GemvLoader<4, 2, 8>, 8> dur=   5.31us dram= 52.1% (  510GB/s rd=   2.7MB) sm=  8.0% issue=  8.0% tensor=  0.0% alu=  7.8% fma=  3.5% occ= 56.4/  83% l2hit= 27.8% regs=44 smem=  5.0K grid=640x256 bankcf=0 stalls=long_scoreboard:38.0,imc_miss:10.8,barrier:7.1,wait:2.2
served_M16   void exl3_gemm_kernel<4, 0, 2, 16, 32, 128, 4, 3> dur=  35.36us dram= 76.2% (  748GB/s rd=  26.5MB) sm= 52.1% issue= 52.1% tensor= 14.6% alu= 50.0% fma= 19.9% occ= 33.1/  33% l2hit= 37.6% regs=107 smem= 91.0K grid=128x512 bankcf=41022 stalls=barrier:1.7,wait:1.3,math_pipe_throttle:1.2,selected:1.0
mma_M16      void ag_mma_prestage_kernel<2, 1>(const void *, const __half dur=   2.34us dram=  8.1% (   78GB/s rd=   0.2MB) sm=  4.8% issue=  4.8% tensor=  0.0% alu= 11.1% fma=  5.2% occ= 16.5/ 100% l2hit= 61.9% regs=24 smem=  1.0K grid=80x256 bankcf=0 stalls=long_scoreboard:5.2,imc_miss:3.5,wait:1.4,selected:1.0
mma_M16      void ag_mma_kernel<4, 2, 2, 1, 1>(const void *, const unsign dur=  36.80us dram= 73.2% (  719GB/s rd=  26.5MB) sm= 32.4% issue= 32.4% tensor= 15.3% alu= 33.3% fma= 13.5% occ= 56.0/  67% l2hit= 51.2% regs=64 smem= 17.0K grid=800x256 bankcf=174411 stalls=long_scoreboard:10.0,wait:1.6,math_pipe_throttle:1.3,barrier:1.1
served_M32   void exl3_gemm_kernel<4, 0, 2, 32, 32, 128, 4, 3> dur=  43.97us dram= 61.6% (  605GB/s rd=  26.6MB) sm= 45.4% issue= 45.4% tensor= 23.3% alu= 42.2% fma= 16.6% occ= 33.3/  33% l2hit= 53.8% regs=112 smem= 91.0K grid=128x512 bankcf=179575 stalls=barrier:2.2,math_pipe_throttle:2.2,wait:1.2,selected:1.0
mma_M32      void ag_mma_prestage_kernel<4, 1>(const void *, const __half dur=   2.46us dram= 14.4% (  141GB/s rd=   0.3MB) sm=  8.8% issue=  8.8% tensor=  0.0% alu= 13.0% fma=  6.1% occ= 20.2/ 100% l2hit= 58.6% regs=24 smem=  1.0K grid=160x256 bankcf=0 stalls=long_scoreboard:6.0,imc_miss:3.1,wait:1.4,selected:1.0
mma_M32      void ag_mma_kernel<4, 2, 4, 1, 1>(const void *, const unsign dur=  43.90us dram= 61.8% (  607GB/s rd=  26.6MB) sm= 30.4% issue= 30.4% tensor= 26.4% alu= 30.5% fma= 12.3% occ= 43.3/  50% l2hit= 67.1% regs=64 smem= 33.0K grid=800x256 bankcf=379658 stalls=long_scoreboard:5.5,math_pipe_throttle:3.0,wait:1.6,barrier:1.0
