Mailing-List: contact dev-help@mahout.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@mahout.apache.org
Date: Mon, 16 Jan 2017 02:28:29 +0000 (UTC)
From: "Andrew Palumbo (JIRA)" <jira@apache.org>
To: dev@mahout.apache.org
Message-ID: <JIRA.12913123.1447546735000.11952.1484533709284@Atlassian.JIRA>
In-Reply-To: <JIRA.12913123.1447546735000@Atlassian.JIRA>
References: <JIRA.12913123.1447546735000@Atlassian.JIRA> <JIRA.12913123.1447546735305@jira-lw-us.apache.org>
Subject: [jira] [Updated] (MAHOUT-1791) Automatic threading for java based
 mmul in the front end and the backend.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Mon, 16 Jan 2017 02:28:47 -0000


     [ https://issues.apache.org/jira/browse/MAHOUT-1791?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Andrew Palumbo updated MAHOUT-1791:
-----------------------------------
    Sprint: Jan/Feb-2017  (was: Jan/Feb-2016)

> Automatic threading for java based mmul in the front end and the backend.
> -------------------------------------------------------------------------
>
>                 Key: MAHOUT-1791
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-1791
>             Project: Mahout
>          Issue Type: Improvement
>    Affects Versions: 0.11.1, 0.12.0, 0.11.2
>            Reporter: Dmitriy Lyubimov
>            Assignee: Andrew Musselman
>            Priority: Critical
>             Fix For: 0.13.0
>
>
> As we know, we are still struggling with decisions which path to take for bare metal accelerations in in-core math. 
> Meanwhile, a simple no-brainer improvement though is to add decision paths and apply multithreaded matrix-matrix multiplication (and maybe even others; but mmul perhaps is the most prominent beneficiary here at the moment which is both easy to do and to have a statistically significant improvement) 
> So multithreaded logic addition to mmul is one path. 
> Another path is automatic adjustment of multithreading. 
> In front end, we probably want to utilize all cores available. 
> in the backend, we can oversubscribe cores but probably doing so by more than 2x or 3x is unadvisable because of point of diminishing returns driven by growing likelihood of context switching overhead.


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)