Skip to content

Commit 9307b01

Browse files
committed
added more metrics to the gpu metadata.csv
1 parent c24a3e8 commit 9307b01

File tree

1 file changed

+7
-0
lines changed

1 file changed

+7
-0
lines changed

gpu/metadata.csv

+7
Original file line numberDiff line numberDiff line change
@@ -29,6 +29,13 @@ gpu.nvlink.throughput.data.rx,gauge,,kilobyte,second,RX data throughput of all N
2929
gpu.nvlink.throughput.data.tx,gauge,,kilobyte,second,TX data throughput of all NVLINK links,0,gpu,nvlink.throughput.data.tx,,
3030
gpu.nvlink.throughput.raw.rx,gauge,,kilobyte,second,RX raw throughput of all NVLINK links,0,gpu,nvlink.throughput.raw.rx,,
3131
gpu.nvlink.throughput.raw.tx,gauge,,kilobyte,second,TX raw throughput of all NVLINK links,0,gpu,nvlink.throughput.raw.tx,,
32+
gpu.nvlink.speed,gauge,,megabyte,second,common speed of all NVLINK links,0,gpu,nvlink.speed,,
33+
gpu.nvlink.nvswitch_connected,gauge,,,,Number of NVLinks connected to the NVSwitch,0,gpu,nvlink.nvswitch_connected,,
34+
gpu.nvlink.errors.crc.data,counter,,,,NVLink data CRC error counter, total for all links,0,gpu,nvlink.errors.crc.data,,
35+
gpu.nvlink.errors.crc.flit,counter,,,,NVLink flow control CRC error counter, total for all links,0,gpu,nvlink.errors.crc.flit,,
36+
gpu.nvlink.errors.ecc,counter,,,,NVLink ECC error counter, total for all links,0,gpu,nvlink.errors.ecc,,
37+
gpu.nvlink.errors.recovery,counter,,,,NVLink recovery error counter, total for all links,0,gpu,nvlink.errors.recovery,,
38+
gpu.nvlink.errors.replay,counter,,,,NVLink replay error counter, total for all links,0,gpu,nvlink.errors.replay,,
3239
gpu.pci.replay_counter,count,,,,PCI replay counter,0,gpu,pci.replay_counter,,
3340
gpu.pci.throughput.rx,gauge,,byte,second,Bytes received through PCI from the GPU device per second,0,gpu,pci.throughput.rx,,
3441
gpu.pci.throughput.tx,gauge,,byte,second,Bytes transmitted through PCI to the GPU device per second,0,gpu,pci.throughput.tx,,

0 commit comments

Comments
 (0)