| 1 |
|
| 2 |
Platform: NVIDIA CUDA |
| 3 |
Device: GeForce MX130 |
| 4 |
Driver version : 455.38 (Linux x64) |
| 5 |
Compute units : 3 |
| 6 |
Clock frequency : 1189 MHz |
| 7 |
|
| 8 |
Global memory bandwidth (GBPS) |
| 9 |
float : 32.25 |
| 10 |
float2 : 33.80 |
| 11 |
float4 : 34.93 |
| 12 |
float8 : 35.33 |
| 13 |
float16 : 25.96 |
| 14 |
|
| 15 |
Single-precision compute (GFLOPS) |
| 16 |
float : 573.05 |
| 17 |
float2 : 845.75 |
| 18 |
float4 : 862.53 |
| 19 |
float8 : 857.37 |
| 20 |
float16 : 854.60 |
| 21 |
|
| 22 |
No half precision support! Skipped |
| 23 |
|
| 24 |
Double-precision compute (GFLOPS) |
| 25 |
double : 27.64 |
| 26 |
double2 : 27.63 |
| 27 |
double4 : 27.58 |
| 28 |
double8 : 27.49 |
| 29 |
double16 : 27.28 |
| 30 |
|
| 31 |
Integer compute (GIOPS) |
| 32 |
int : 261.52 |
| 33 |
int2 : 290.67 |
| 34 |
int4 : 293.02 |
| 35 |
int8 : 278.32 |
| 36 |
int16 : 267.97 |
| 37 |
|
| 38 |
Integer compute Fast 24bit (GIOPS) |
| 39 |
int : 261.52 |
| 40 |
int2 : 290.59 |
| 41 |
int4 : 293.09 |
| 42 |
int8 : 291.33 |
| 43 |
int16 : 289.98 |
| 44 |
|
| 45 |
Transfer bandwidth (GBPS) |
| 46 |
enqueueWriteBuffer : 2.81 |
| 47 |
enqueueReadBuffer : 3.23 |
| 48 |
enqueueWriteBuffer non-blocking : 2.84 |
| 49 |
enqueueReadBuffer non-blocking : 3.09 |
| 50 |
enqueueMapBuffer(for read) : 3.12 |
| 51 |
memcpy from mapped ptr : 7.80 |
| 52 |
enqueueUnmap(after write) : 3.11 |
| 53 |
memcpy to mapped ptr : 7.89 |
| 54 |
|
| 55 |
Kernel launch latency : 7.62 us |
| 56 |
|