Send hpx-devel mailing list submissions to
hpx-devel@stellar.cct.lsu.edu
To subscribe or unsubscribe via the World Wide Web, visit
https://mail.cct.lsu.edu/mailman/listinfo/hpx-devel
or, via email, send a message with subject or body 'help' to
hpx-devel-requ...@stellar.cct.lsu.edu
You can reach the person managing the list at
hpx-devel-ow...@stellar.cct.lsu.edu
When replying, please edit your Subject line so it is more specific
than "Re: Contents of hpx-devel digest..."
Today's Topics:
1. Fwd: Question about GSoC project (Patrick Diehl)
----------------------------------------------------------------------
Message: 1
Date: Sat, 10 Mar 2018 07:47:14 -0500
From: Patrick Diehl <patrickdie...@gmail.com>
Subject: [hpx-devel] Fwd: Question about GSoC project
To: hpx-devel@stellar.cct.lsu.edu
Message-ID: <0d617434-e2c3-15f6-e317-f987d8a7d...@gmail.com>
Content-Type: text/plain; charset="utf-8"
-------- Forwarded Message --------
Subject: Question about GSoC project
Date: Sat, 10 Mar 2018 17:11:51 +0900
From: ??? <msc...@naver.com>
Reply-To: ??? <msc...@naver.com>
To: Patrick Diehl <patrickdie...@gmail.com>
Hi Patrick,
I was thinking about the features I could use for making a prediction.
While current HPX constructs such as persistent_auto_chunk_size use the
mean runtime of 1% of the tasks,
I think having more statistical information about the 1% runtime could
be beneficial.
For example, having information about the min and max runtime, variance,
etc.. .
Imagine a situation where the mean runtime and the max runtime differ a lot.
Then choosing a finer chunk size could lead to better performance since
the tasks that spike in runtime would act less as bottlenecks.
?
Though implementing this would require to modify or specialize
get_bulk_iteration_shape function.
Would this idea be appropriate to go into the proposal?
?
Thanks for reading.
Ray Kim.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
Url :
http://mail.cct.lsu.edu/pipermail/hpx-devel/attachments/20180310/70ca2ef1/attachment-0001.bin
------------------------------
_______________________________________________
hpx-devel mailing list
hpx-devel@stellar.cct.lsu.edu
https://mail.cct.lsu.edu/mailman/listinfo/hpx-devel
End of hpx-devel Digest, Vol 38, Issue 1
****************************************
