Debian Codegraph @

results/NVIDIA_CUDA/GeForce_RTX_2080_Super.log

raw
1
2 Platform: NVIDIA CUDA
3 Device: GeForce RTX 2080 SUPER
4 Driver version : 440.64 (Linux x64)
5 Compute units : 48
6 Clock frequency : 1830 MHz
7
8 Global memory bandwidth (GBPS)
9 float : 379.44
10 float2 : 400.04
11 float4 : 406.76
12 float8 : 421.57
13 float16 : 387.33
14
15 Single-precision compute (GFLOPS)
16 float : 11928.26
17 float2 : 11929.92
18 float4 : 11826.57
19 float8 : 11323.94
20 float16 : 11032.11
21
22 No half precision support! Skipped
23
24 Double-precision compute (GFLOPS)
25 double : 368.15
26 double2 : 362.51
27 double4 : 361.13
28 double8 : 350.26
29 double16 : 356.73
30
31 Integer compute (GIOPS)
32 int : 10998.23
33 int2 : 11085.42
34 int4 : 11515.90
35 int8 : 11564.03
36 int16 : 11580.59
37
38 Transfer bandwidth (GBPS)
39 enqueueWriteBuffer : 9.02
40 enqueueReadBuffer : 8.65
41 enqueueMapBuffer(for read) : 11.40
42 memcpy from mapped ptr : 11.83
43 enqueueUnmap(after write) : 12.93
44 memcpy to mapped ptr : 11.99
45
46 Kernel launch latency : 3.46 us
47