Objet : Developers list for StarPU
Archives de la liste
- From: Alfredo Buttari <alfredo.buttari@enseeiht.fr>
- To: Samuel Thibault <samuel.thibault@inria.fr>, Alfredo Buttari <alfredo.buttari@enseeiht.fr>, starpu-devel@lists.gforge.inria.fr, Emmanuel Agullo <Emmanuel.Agullo@inria.fr>, Abdou Guermouche <abdou.guermouche@labri.fr>
- Subject: Re: [Starpu-devel] huge regression in LWS scheduler
- Date: Thu, 9 Jun 2016 09:51:27 +0200
- Authentication-results: mail3-smtp-sop.national.inria.fr; spf=None smtp.pra=alfredo.buttari@gmail.com; spf=Pass smtp.mailfrom=alfredo.buttari@gmail.com; spf=None smtp.helo=postmaster@mail-wm0-f48.google.com
- Ironport-phdr: 9a23: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
- List-archive: <http://lists.gforge.inria.fr/pipermail/starpu-devel/>
- List-id: "Developers list. For discussion of new features, code changes, etc." <starpu-devel.lists.gforge.inria.fr>
Hello,
I cannot use that machine currently, but I have managed to reproduce the problem (to a lower extent) on another smaller (24 cores) system (with a smaller dataset). Here are the results for a sequence of 8 runs in each configuration:
| trunk eager | trunk lws | branch-1.2 eager | branch-1.2 lws |
| 542 | 518 | 536 | 540 |
| 540 | 381 | 535 | 539 |
| 540 | 522 | 533 | 540 |
| 536 | 275 | 536 | 537 |
| 542 | 410 | 537 | 537 |
| 541 | 269 | 533 | 542 |
| 540 | 518 | 535 | 542 |
| 538 | 382 | 530 | 540 |
all the results are very consistent except in the "trunk lws" column where results are very variable and never as good as in the other configurations.
Ciao
Alfredo
On Wed, Jun 8, 2016 at 6:58 PM, Samuel Thibault <samuel.thibault@inria.fr> wrote:
Alfredo Buttari, on Wed 08 Jun 2016 15:20:01 +0200, wrote:
> starpu-1.2.0rc5 eager : 513 Gflop/s
> starpu-1.2.0rc5 lws : 623 Gflop/s
Could you try the 1.2 branch?
-----------------------------------------
Alfredo Buttari, PhD
CNRS-IRIT
2 rue Camichel, 31071 Toulouse, France
http://buttari.perso.enseeiht.fr
Alfredo Buttari, PhD
CNRS-IRIT
2 rue Camichel, 31071 Toulouse, France
http://buttari.perso.enseeiht.fr
- [Starpu-devel] huge regression in LWS scheduler, Alfredo Buttari, 08/06/2016
- Re: [Starpu-devel] huge regression in LWS scheduler, Samuel Thibault, 08/06/2016
- Re: [Starpu-devel] huge regression in LWS scheduler, Alfredo Buttari, 09/06/2016
- Re: [Starpu-devel] huge regression in LWS scheduler, Samuel Thibault, 08/06/2016
Archives gérées par MHonArc 2.6.19+.