• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1
2Platform: AMD Accelerated Parallel Processing
3  Device: Spectre
4    Driver version  : 1573.4 (VM) (Linux x64)
5    Compute units   : 8
6    Clock frequency : 720 MHz
7
8    Global memory bandwidth (GBPS)
9      float   : 27.21
10      float2  : 28.08
11      float4  : 22.67
12      float8  : 15.10
13      float16 : 8.28
14
15    Single-precision compute (GFLOPS)
16      float   : 733.13
17      float2  : 732.25
18      float4  : 730.70
19      float8  : 726.86
20      float16 : 718.54
21
22    Double-precision compute (GFLOPS)
23      double   : 45.98
24      double2  : 45.95
25      double4  : 45.91
26      double8  : 45.81
27      double16 : 45.61
28
29    Integer compute (GIOPS)
30      int   : 146.82
31      int2  : 146.91
32      int4  : 146.03
33      int8  : 145.64
34      int16 : 146.70
35
36    Transfer bandwidth (GBPS)
37      enqueueWriteBuffer         : 8.42
38      enqueueReadBuffer          : 5.40
39      enqueueMapBuffer(for read) : 4357.72
40        memcpy from mapped ptr   : 3.76
41      enqueueUnmap(after write)  : 8918.12
42        memcpy to mapped ptr     : 3.54
43
44    Kernel launch latency : 33.79 us
45
46  Device: AMD A10-7850K APU with Radeon(TM) R7 Graphics
47    Driver version  : 1573.4 (sse2,avx,fma4) (Linux x64)
48    Compute units   : 4
49    Clock frequency : 4000 MHz
50
51    Global memory bandwidth (GBPS)
52      float   : 9.72
53      float2  : 10.97
54      float4  : 12.09
55      float8  : 11.17
56      float16 : 14.40
57
58    Single-precision compute (GFLOPS)
59      float   : 3.12
60      float2  : 5.69
61      float4  : 12.62
62      float8  : 22.69
63      float16 : 44.04
64
65    Double-precision compute (GFLOPS)
66      double   : 3.11
67      double2  : 6.28
68      double4  : 11.28
69      double8  : 23.00
70      double16 : 28.83
71
72    Integer compute (GIOPS)
73      int   : 6.16
74      int2  : 4.10
75      int4  : 13.76
76      int8  : 26.61
77      int16 : 30.75
78
79    Transfer bandwidth (GBPS)
80      enqueueWriteBuffer         : 8.31
81      enqueueReadBuffer          : 5.45
82      enqueueMapBuffer(for read) : 10324.44
83        memcpy from mapped ptr   : 3.76
84      enqueueUnmap(after write)  : 19311.90
85        memcpy to mapped ptr     : 3.56
86
87    Kernel launch latency : 11.89 us
88
89