singa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Muhammad Hamdan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SINGA-308) CPU-GPU parallelism
Date Tue, 11 Apr 2017 06:49:41 GMT

    [ https://issues.apache.org/jira/browse/SINGA-308?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15963915#comment-15963915
] 

Muhammad Hamdan commented on SINGA-308:
---------------------------------------

Thank you! that worked. 
However, it  seemed pretty much like the CPU based implementation/training. CPU utilization
was 100% while GPU was about 9% MAX. 
I thought this implementation would do the opposite and would rely on the CPU to work along
the GPU to speed up the process comparably to a single GPU.

> CPU-GPU parallelism 
> --------------------
>
>                 Key: SINGA-308
>                 URL: https://issues.apache.org/jira/browse/SINGA-308
>             Project: Singa
>          Issue Type: Test
>          Components: Core, PySINGA
>         Environment: Ubuntu 16.04
> CPU-GPU of the same machine
>            Reporter: Muhammad Hamdan
>              Labels: test
>
> Is it possible to parallelize the alexnet model for the cifar10 example on a CPU and
GPU instead of 2-GPUs ? Assuming asynchronous communication between the two components



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message