[
https://issues.apache.org/jira/browse/LUCENE-7745?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15945065#comment-15945065
]
David Smiley commented on LUCENE-7745:
--------------------------------------
vikash: not all working code contributed to any open-source project is
necessarily welcome. Usually it is but sometimes project members or ASF rules
insist on certain things for the perceived greater good. In this case, I
believe Uwe doesn't want Lucene to include anything that would only work with
certain hardware or JVM vendors -- even if it was optional opt-in. If
hypothetically nobody had such concerns here, be aware that any 3rd party
(non-ASF) libraries need to meet certain qualifications. For example, *if*
whatever Java CUDA library you find happens to be licensed as GPL, then it's
incompatible with ASF run projects like this one. That's a hypothetical; I
have no idea what Java CUDA libraries exist and what their licenses are.
Regardless... if you come up with something useful, it's probably not necessary
that Lucene itself change, and as seen here we have some willingness to change
Lucene (details TBD) if it enables people to use Lucene with CUDA. Lucene has
many extension points already. Though I could imagine you might unfortunately
need to copy/fork some long source files -- Uwe mentioned some.
Good luck.
> Explore GPU acceleration
> ------------------------
>
> Key: LUCENE-7745
> URL: https://issues.apache.org/jira/browse/LUCENE-7745
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Ishan Chattopadhyaya
> Labels: gsoc2017, mentor
>
> There are parts of Lucene that can potentially be speeded up if computations
> were to be offloaded from CPU to the GPU(s). With commodity GPUs having as
> high as 12GB of high bandwidth RAM, we might be able to leverage GPUs to
> speed parts of Lucene (indexing, search).
> First that comes to mind is spatial filtering, which is traditionally known
> to be a good candidate for GPU based speedup (esp. when complex polygons are
> involved). In the past, Mike McCandless has mentioned that "both initial
> indexing and merging are CPU/IO intensive, but they are very amenable to
> soaking up the hardware's concurrency."
> I'm opening this issue as an exploratory task, suitable for a GSoC project. I
> volunteer to mentor any GSoC student willing to work on this this summer.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
