Project

General

Profile

MulticoreDiscussions

These pages are no longer relevant, go to [[cet-multicore:CET Multicore Project]] for the up-to-date information.

6/4/2010

What are we doing with regards to CSR work?
Currently we are going to try to make small steps forward understanding where we shall go
given the limited time we have to work on this.

Over the next three months to six months, we will be tackling several areas:

1. Training and conferences - This includes the summer cluster computing conference,
Supercomputing, other multicore classes that come up.

2. Programming Techniques and Tools - Pick from the available programming languages
for multicore / HPC computing (not C, Java, C++) and prepare a short talk that discusses
possible uses for that language in our community and what is interesting about it or
unique about it regarding concurrency. Tell how it might benefit the community.

3. External Interactions - Find an external group with experience in multicore / HPC
programming that may be applicable to our needs and arrange to meet with them
(perhaps Argonne, UC, UIUC, NWU, IIT) to learn more about what they do and to know
who is involved.

4. Application identification - identify a part of any application that might benefit
from multicore programming techniques and HPC cluster use and report on how that might
be accomplished. Arrange for talks from they groups to learn about their problems.

5. Work with Paolo Calafura of ATLAS to use the PTU measurements in a meaningful way.
Also work with CMS to do something constructive.

6. Think about total problem optimization involved detector design exploration, trigger
studies, background studies.

2010.7.28

For the remainder of summer, number 5 above might be an interesting choice. This needs to
be a low-level activity.

The other thing that needs attention is the LQCD project. There is time to be spent on things here.

Notes