Skip to content
Snippets Groups Projects
Commit 5f7751f2 authored by Milan Jaros's avatar Milan Jaros
Browse files

Update README.md

parent ae330664
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
Showing
with 38 additions and 37 deletions
README.md
+ 38
− 37
View file @ 5f7751f2
...@@ -29,47 +29,48 @@ Tests (sendrecv_perf): 2x nodes - 2x GPUs ...@@ -29,47 +29,48 @@ Tests (sendrecv_perf): 2x nodes - 2x GPUs
```shell ```shell
# Using devices # Using devices
# Rank 0 Pid 1003 on acn11 device 0 [0x48] NVIDIA A100-SXM4-40GB # Rank 0 Pid 48277 on acn18 device 0 [0x48] NVIDIA A100-SXM4-40GB
# Rank 1 Pid 14510 on acn42 device 0 [0x48] NVIDIA A100-SXM4-40GB # Rank 1 Pid 114173 on acn29 device 0 [0x48] NVIDIA A100-SXM4-40GB
# #
# out-of-place in-place # out-of-place in-place
# size count type time algbw busbw error time algbw busbw error # size count type time algbw busbw error time algbw busbw error
# (B) (elements) (us) (GB/s) (GB/s) (us) (GB/s) (GB/s) # (B) (elements) (us) (GB/s) (GB/s) (us) (GB/s) (GB/s)
0 0 float 27.61 0.00 0.00 0e+00 27.25 0.00 0.00 0e+00 0 0 float 27.57 0.00 0.00 0e+00 27.52 0.00 0.00 0e+00
0 0 float 27.51 0.00 0.00 0e+00 27.47 0.00 0.00 0e+00 0 0 float 27.26 0.00 0.00 0e+00 26.89 0.00 0.00 0e+00
4 1 float 37.13 0.00 0.00 0e+00 36.89 0.00 0.00 0e+00 4 1 float 36.95 0.00 0.00 0e+00 37.04 0.00 0.00 0e+00
8 2 float 37.08 0.00 0.00 0e+00 37.32 0.00 0.00 0e+00 8 2 float 37.05 0.00 0.00 0e+00 36.84 0.00 0.00 0e+00
16 4 float 36.89 0.00 0.00 0e+00 36.52 0.00 0.00 0e+00 16 4 float 37.01 0.00 0.00 0e+00 36.73 0.00 0.00 0e+00
32 8 float 36.95 0.00 0.00 0e+00 36.84 0.00 0.00 0e+00 32 8 float 36.77 0.00 0.00 0e+00 36.54 0.00 0.00 0e+00
64 16 float 36.82 0.00 0.00 0e+00 36.63 0.00 0.00 0e+00 64 16 float 36.65 0.00 0.00 0e+00 36.75 0.00 0.00 0e+00
128 32 float 36.87 0.00 0.00 0e+00 36.89 0.00 0.00 0e+00 128 32 float 36.88 0.00 0.00 0e+00 36.51 0.00 0.00 0e+00
256 64 float 36.82 0.01 0.01 0e+00 36.85 0.01 0.01 0e+00 256 64 float 36.62 0.01 0.01 0e+00 37.23 0.01 0.01 0e+00
512 128 float 36.86 0.01 0.01 0e+00 37.14 0.01 0.01 0e+00 512 128 float 36.77 0.01 0.01 0e+00 36.69 0.01 0.01 0e+00
1024 256 float 37.26 0.03 0.03 0e+00 37.51 0.03 0.03 0e+00 1024 256 float 36.74 0.03 0.03 0e+00 37.16 0.03 0.03 0e+00
2048 512 float 37.46 0.05 0.05 0e+00 38.09 0.05 0.05 0e+00 2048 512 float 37.24 0.05 0.05 0e+00 37.16 0.06 0.06 0e+00
4096 1024 float 38.06 0.11 0.11 0e+00 37.92 0.11 0.11 0e+00 4096 1024 float 38.00 0.11 0.11 0e+00 37.92 0.11 0.11 0e+00
8192 2048 float 38.95 0.21 0.21 0e+00 38.57 0.21 0.21 0e+00 8192 2048 float 38.76 0.21 0.21 0e+00 39.76 0.21 0.21 0e+00
16384 4096 float 39.68 0.41 0.41 0e+00 39.70 0.41 0.41 0e+00 16384 4096 float 40.50 0.40 0.40 0e+00 40.16 0.41 0.41 0e+00
32768 8192 float 43.73 0.75 0.75 0e+00 43.66 0.75 0.75 0e+00 32768 8192 float 43.42 0.75 0.75 0e+00 43.40 0.75 0.75 0e+00
65536 16384 float 54.43 1.20 1.20 0e+00 54.10 1.21 1.21 0e+00 65536 16384 float 52.10 1.26 1.26 0e+00 50.81 1.29 1.29 0e+00
131072 32768 float 68.75 1.91 1.91 0e+00 69.91 1.87 1.87 0e+00 131072 32768 float 63.40 2.07 2.07 0e+00 64.73 2.03 2.03 0e+00
262144 65536 float 96.53 2.72 2.72 0e+00 96.38 2.72 2.72 0e+00 262144 65536 float 85.99 3.05 3.05 0e+00 87.69 2.99 2.99 0e+00
524288 131072 float 148.8 3.52 3.52 0e+00 149.6 3.50 3.50 0e+00 524288 131072 float 127.9 4.10 4.10 0e+00 127.3 4.12 4.12 0e+00
1048576 262144 float 239.6 4.38 4.38 0e+00 237.5 4.42 4.42 0e+00 1048576 262144 float 205.6 5.10 5.10 0e+00 199.9 5.24 5.24 0e+00
2097152 524288 float 429.8 4.88 4.88 0e+00 427.2 4.91 4.91 0e+00 2097152 524288 float 338.1 6.20 6.20 0e+00 343.9 6.10 6.10 0e+00
4194304 1048576 float 763.0 5.50 5.50 0e+00 882.2 4.75 4.75 0e+00 4194304 1048576 float 585.6 7.16 7.16 0e+00 592.0 7.09 7.09 0e+00
8388608 2097152 float 1520.9 5.52 5.52 0e+00 1737.4 4.83 4.83 1e+00 8388608 2097152 float 1128.5 7.43 7.43 0e+00 1134.9 7.39 7.39 0e+00
16777216 4194304 float 3235.5 5.19 5.19 0e+00 3247.8 5.17 5.17 1e+00 16777216 4194304 float 2213.3 7.58 7.58 0e+00 2221.3 7.55 7.55 0e+00
33554432 8388608 float 6090.4 5.51 5.51 0e+00 6082.5 5.52 5.52 1e+00 33554432 8388608 float 4378.5 7.66 7.66 0e+00 4381.0 7.66 7.66 0e+00
67108864 16777216 float 12354 5.43 5.43 0e+00 12384 5.42 5.42 1e+00 67108864 16777216 float 8689.8 7.72 7.72 0e+00 8688.1 7.72 7.72 0e+00
134217728 33554432 float 24488 5.48 5.48 0e+00 24472 5.48 5.48 1e+00 134217728 33554432 float 17334 7.74 7.74 0e+00 17343 7.74 7.74 0e+00
268435456 67108864 float 48557 5.53 5.53 0e+00 48579 5.53 5.53 1e+00 268435456 67108864 float 34589 7.76 7.76 0e+00 34583 7.76 7.76 1e+00
536870912 134217728 float 97378 5.51 5.51 0e+00 97041 5.53 5.53 1e+00 536870912 134217728 float 69103 7.77 7.77 0e+00 69119 7.77 7.77 1e+00
1073741824 268435456 float 194355 5.52 5.52 0e+00 194175 5.53 5.53 1e+00 1073741824 268435456 float 138171 7.77 7.77 0e+00 138151 7.77 7.77 1e+00
2147483648 536870912 float 388342 5.53 5.53 0e+00 388345 5.53 5.53 1e+00 2147483648 536870912 float 276503 7.77 7.77 0e+00 276245 7.77 7.77 1e+00
4294967296 1073741824 float 776666 5.53 5.53 0e+00 776807 5.53 5.53 1e+00 4294967296 1073741824 float 552562 7.77 7.77 0e+00 552876 7.77 7.77 1e+00
8589934592 2147483648 float 1106176 7.77 7.77 0e+00 1105161 7.77 7.77 1e+00
# Out of bounds values : 0 OK # Out of bounds values : 0 OK
# Avg bus bandwidth : 2.41644 # Avg bus bandwidth : 3.38821
# #
``` ```
...@@ -603,4 +604,4 @@ bandwidth: host to host ...@@ -603,4 +604,4 @@ bandwidth: host to host
1048576 22964.75 1048576 22964.75
2097152 22973.91 2097152 22973.91
4194304 23635.41 4194304 23635.41
``` ```
\ No newline at end of file
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment