[jira] [Closed] (SYSTEMML-320) Performance features for sparse rdd operations

["Sebastian Baunsgaard (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/SYSTEMML-320?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sebastian Baunsgaard closed SYSTEMML-320.
-----------------------------------------
    Fix Version/s: SystemDS 2.0
       Resolution: Fixed

Should no longer be the case

> Performance features for sparse rdd operations
> ----------------------------------------------
>
>                 Key: SYSTEMML-320
>                 URL: https://issues.apache.org/jira/browse/SYSTEMML-320
>             Project: SystemDS
>          Issue Type: Task
>            Reporter: Matthias Boehm
>            Assignee: Matthias Boehm
>            Priority: Blocker
>             Fix For: SystemDS 2.0
>
>
> The spark backend in its current state has various performance issues on sparse and especially ultra-sparse datasets. This task tracks all related performance features such as
> * Decisions on serialized rdd storage, incl performance improvements serialization
> * Per-block overhead reduction for rdd operations
> * New ultra-sparse in-memory block representation, incl operations over it



--
This message was sent by Atlassian Jira
(v8.3.4#803005)