You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@mahout.apache.org by "Trevor Grant (Jira)" <ji...@apache.org> on 2024/01/31 21:43:00 UTC
[jira] [Resolved] (MAHOUT-2079) investigat Cuda as a backend for in core matricies.
[ https://issues.apache.org/jira/browse/MAHOUT-2079?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Trevor Grant resolved MAHOUT-2079.
----------------------------------
Resolution: Won't Fix
Jira Cleanup 1/31/24
> investigat Cuda as a backend for in core matricies.
> ---------------------------------------------------
>
> Key: MAHOUT-2079
> URL: https://issues.apache.org/jira/browse/MAHOUT-2079
> Project: Mahout
> Issue Type: Wish
> Affects Versions: 0.14.1
> Reporter: Andrew Palumbo
> Assignee: Andrew Palumbo
> Priority: Major
> Fix For: 14.2
>
>
> with nvidia card memories now exceeding what used to be max for some machines.. 16 G easily it would be good to be able to dumb into Cuda memory directly from a statment or assignment. look into backing in-core matrices with CIDA memory or CUDA Shared memory.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)