2019-07-04 03:51:46 |
Jerry Kao |
bug |
|
|
added bug |
2019-07-04 03:51:56 |
Jerry Kao |
tags |
|
ce-qa-concern |
|
2019-07-04 03:52:03 |
Jerry Kao |
plainbox-provider-checkbox: status |
New |
Confirmed |
|
2019-07-04 03:52:05 |
Jerry Kao |
plainbox-provider-checkbox: importance |
Undecided |
High |
|
2019-07-04 03:52:27 |
Jerry Kao |
plainbox-provider-checkbox: assignee |
|
Ray Chen (ray.chen) |
|
2019-07-04 03:53:38 |
Jerry Kao |
description |
NVLink is more and more visible on AI/ML workstations. We need a job or a suit of test cases to verify NVLink.
Currently, QA are checking NVLink by command nvidia-smi nvlink -c
Then install cuda for bandwidth testing (detail please refer to https://docs.google.com/document/d/1WuhegvZMrghhnHbbXTv1sX8GtP4U8GnrGRlbgEC_GVA/edit)
$ git clone https://github.com/NVIDIA/cuda-samples.git
$ cd cuda-samples/Samples/p2pBandwidthLatencyTest
$ make
$ sudo ./p2pBandwidthLatencyTest
Tensorflow is a 3rd party application for AI/ML. It's not necessary to test (with) Tensorflow. |
NVLink is more and more visible on AI/ML workstations. We need a job or a suit of test cases to verify NVLink.
Currently, QA are checking NVLink by command nvidia-smi nvlink -c
Then install cuda for bandwidth testing (detail please refer to https://docs.google.com/document/d/1WuhegvZMrghhnHbbXTv1sX8GtP4U8GnrGRlbgEC_GVA/edit)
$ git clone https://github.com/NVIDIA/cuda-samples.git
$ cd cuda-samples/Samples/p2pBandwidthLatencyTest
$ make
$ sudo ./p2pBandwidthLatencyTest
Tensorflow is a 3rd party application for AI/ML. It's not necessary to test (with) Tensorflow.
The job needs to depend on NVidia GPU and NVidia driver. |
|
2019-07-04 03:54:32 |
Jerry Kao |
description |
NVLink is more and more visible on AI/ML workstations. We need a job or a suit of test cases to verify NVLink.
Currently, QA are checking NVLink by command nvidia-smi nvlink -c
Then install cuda for bandwidth testing (detail please refer to https://docs.google.com/document/d/1WuhegvZMrghhnHbbXTv1sX8GtP4U8GnrGRlbgEC_GVA/edit)
$ git clone https://github.com/NVIDIA/cuda-samples.git
$ cd cuda-samples/Samples/p2pBandwidthLatencyTest
$ make
$ sudo ./p2pBandwidthLatencyTest
Tensorflow is a 3rd party application for AI/ML. It's not necessary to test (with) Tensorflow.
The job needs to depend on NVidia GPU and NVidia driver. |
NVLink is more and more visible on AI/ML workstations. We need a job or a suit of test cases to verify NVLink.
Currently, QA are checking NVLink support by command nvidia-smi nvlink -c
Then install cuda for bandwidth testing (detail please refer to https://docs.google.com/document/d/1WuhegvZMrghhnHbbXTv1sX8GtP4U8GnrGRlbgEC_GVA/edit)
$ git clone https://github.com/NVIDIA/cuda-samples.git
$ cd cuda-samples/Samples/p2pBandwidthLatencyTest
$ make
$ sudo ./p2pBandwidthLatencyTest
Tensorflow is a 3rd party application for AI/ML. It's not necessary to test (with) Tensorflow.
The job needs to depend on NVidia GPU and NVidia driver. |
|
2019-10-03 13:19:30 |
Sylvain Pineau |
merge proposal linked |
|
https://code.launchpad.net/~ray.chen/plainbox-provider-checkbox/+git/plainbox-provider-checkbox/+merge/373544 |
|
2019-10-06 08:33:03 |
Sylvain Pineau |
plainbox-provider-checkbox: milestone |
|
0.50.0 |
|
2019-10-18 08:26:32 |
Sylvain Pineau |
plainbox-provider-checkbox: status |
Confirmed |
In Progress |
|
2019-10-18 08:51:35 |
Sylvain Pineau |
plainbox-provider-checkbox: status |
In Progress |
Fix Committed |
|
2019-11-14 09:03:57 |
Ray Chen |
tags |
ce-qa-concern |
ce-qa-concern cqa-verified |
|
2019-11-15 09:34:10 |
Devices Certification Bot |
plainbox-provider-checkbox: status |
Fix Committed |
Fix Released |
|