June 2011 - Imudi-pc - mailman3.cels.anl.gov

nominations for invited speaker / panel members
by Dries Kimpe 02 Jul '11

02 Jul '11

Hi, if anybody has nomimations for people that could serve on the discussion panel, or could give an invited talk, please reply *to this message* (make sure to respond to the list). I'll compile all suggestions on Thursday, and setup a doodle. At that point I'll send out another email requesting everybody to cast a vote for their preferred panel members/speakers. Names that were mentioned during the concall (and which I still remember): - Torsten Hoefler - Quincey Koziol - Rusty Lusk / Steve Pieper (Asynchronous Dynamic Load Balancing library) Other interesting people (probably also mentioned during the call): - system administrators having experience managing a large MPI cluster - Hardware/embedded software designers (especially if they design hardware specifically with MPI features in mind (hardware collectives,...)) - Other MPI libraries: petsc, trilinos Thanks, Dries

4 6

[[email protected]: Re: [Imudi2011] paper notification]
by Dries Kimpe 28 Jun '11

28 Jun '11

2 1

[[email protected]: Re: [Imudi2011] paper notification]
by Dries Kimpe 28 Jun '11

28 Jun '11

This is about the discrete event simulator paper (see attachment for the paper and the reviews). We discussed this paper during the call, as it had mixed reviews. In the end, we agreed to reject, based on the fact that it was out of scope for the session and that there seemed to be serious issues with the paper. I'm inclined to send a response explaining the review process and clarifying our decision (out of scope, serious issues) and that the decision is final, unless the majority of the PC feels we should reconsider and allow a rebuttal. Greetings, Dries

3 2

Paper #6 (double submission)
by Dries Kimpe 27 Jun '11

27 Jun '11

Paper #6 (which we accepted but which was also submitted to the main conference) was accepted at the main EuroMPI conference. The EuroMPI program co-chairs however want us to take over the paper. (including publication/camera ready). I'll inform the authors. Dries

1 0

Panel/invited talk compensation
by Dries Kimpe 27 Jun '11

27 Jun '11

I checked with the EuroMPI organizers, and there is no possibility to compensate panel members/invited speakers, due to the financial situation of the organisation right now. I think we expected this, but just wanted to confirm. Greetings, Dries

1 0

[[email protected]: Fwd: Re: Paper review results -- paper #3]
by Dries Kimpe 24 Jun '11

24 Jun '11

4 4

paper #3
by Dries Kimpe 24 Jun '11

24 Jun '11

Paper attached.

1 0

reviews for #3
by Dries Kimpe 24 Jun '11

24 Jun '11

Attached are reviews for #3.

1 0

paper #5
by Dries Kimpe 24 Jun '11

24 Jun '11

Attached review of #5, paper. Dries

1 0

Paper review results
by Dries Kimpe 24 Jun '11

24 Jun '11

Good afternoon, We've setup a mailman list to make discussion of the results of the IMUDI paper reviews easier. All PC members were auto-subscribed to the list. We've also setup a conference call for tomorrow morning (10am CDT) for anybody that would like to discuss the results of the papers on the phone instead of via email. The coordinates for that call are: Date and Time: 6/24/2011 @ 10am CDT Phone Number: 570-639-6055 Passcode: 379562# We've attached the initial ranking of the paper reviews below. There are a couple of missing reviews. If you haven't sent us your review(s) yet, and won't be able to do so by early tomorrow morning (8am CDT), please let us know so that we can ensure that all papers have the proper number of reviews (3). RANK Paper # Title 1 (#2) Noncollective Communicator Creation in MPI 2 (#4) Leveraging C++ Meta-Programming 3 (#6) Evaluation of Interpreted Languages with Open MPI 4 (#5) Towards Software-level Virtual Experimentation of MPI-based Million-way Concurrency 5 (#3) A Lightweight C++ Interface to MPI 6 (#1) A Transformation in MPI for topology-aware mapping from Process Topology to Network Topology Currently, it looks like we'll accept the first 3-4 papers. Papers #2,#4 and #6 have unanimous accept. Paper #1 has unanimous reject. Paper #3 was reviewed as reject and two times weak accept. Paper #5 has mixed reviews (definite accept, reject) but is still missing a review. I've attached paper #5 for those who want to take a closer look at the paper. We'll send out an update as soon as we obtain the 3rd review for paper #5. Thanks, Jason Cope and Dries Kimpe

2 1