hi,
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
What we could start working on in the mean time is to get a SIG together, and formulating intent / direction / mechanism etc.
Also, HPC itself is quite a vast subject - and perhaps we should not limit ourselves only to the product base that is being released upstream. So if we were to propose a HPC SIG, are there a few people here who would like to join in the effort and help ?
- KB
On Thu, Oct 30, 2008 at 07:26:57PM +0000, Karanbir Singh wrote:
hi,
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
I haven't heard anything after Tsai Li Ming ltsai@osgdc.org's email "HPC Channel in CentOS" at the beginning of the month.
What we could start working on in the mean time is to get a SIG together, and formulating intent / direction / mechanism etc.
Also, HPC itself is quite a vast subject - and perhaps we should not limit ourselves only to the product base that is being released upstream. So if we were to propose a HPC SIG, are there a few people here who would like to join in the effort and help ?
me :)
Cheers,
Tru
On Thu, 30 Oct 2008, Karanbir Singh wrote:
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
full URL please and I will cause an inquiry from a person i the industry to occur.
- Russ herrold
On Thu, Oct 30, 2008 at 05:01:25PM -0400, R P Herrold wrote:
On Thu, 30 Oct 2008, Karanbir Singh wrote:
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
full URL please and I will cause an inquiry from a person i the industry to occur.
http://www.redhat.com/docs/en-US/hpc/1.0/html/ http://www.redhat.com/hpc/index.html
Cheers,
Tru
Tru Huynh wrote:
On Thu, Oct 30, 2008 at 05:01:25PM -0400, R P Herrold wrote:
On Thu, 30 Oct 2008, Karanbir Singh wrote:
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
full URL please and I will cause an inquiry from a person i the industry to occur.
http://www.redhat.com/docs/en-US/hpc/1.0/html/ http://www.redhat.com/hpc/index.html
Cheers,
Tru
We are also waiting for the SRPMS to be available from Red Hat . I will check with Red Hat to see whether when they will be available.
-Liming
Hi
We would kindly join in and assist.
Our team developed Project Kusu (http://www.hpccommunity.org/kusu) which is the foundation for the Red Hat HPC offering. As mentioned by KB, HPC is very wide, and we definitely need more help and expertise.
For a complete HPC stack the following components would be required:
Infrastructure: 1. OS (centOS) 2. cluster management (kusu) 3. batch scheduler (Lava already included in kusu, SGE, Torque would also be good additions) 4. cluster filesystems (Lustre, GPFS, GFS, pNFS) 5. compilers and libraries (gcc, intel/portland compilers, mpich, openmpi, lapack, scalapack, atlas etc)
Open source HPC applications Kits in 1. bio/life science 2. chemistry and materials science 3. CFD 4. rendering applications 5. financial computations apps 6. numerical routines libraries.... 7. etc
In Kusu we try to provide most of the infrastructure pieces, the next stage of development is to target applications, so that engineers and scientist using Kusu can easily install a cluster with their applications without worrying about the cluster software infrastructure.
Cheers!
Laurence www.HPCCommunity.org
On 31 Oct 2008, at 3:26 AM, Karanbir Singh wrote:
hi,
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
What we could start working on in the mean time is to get a SIG together, and formulating intent / direction / mechanism etc.
Also, HPC itself is quite a vast subject - and perhaps we should not limit ourselves only to the product base that is being released upstream. So if we were to propose a HPC SIG, are there a few people here who would like to join in the effort and help ?
- KB
-- Karanbir Singh : http://www.karan.org/ : 2522219@icq _______________________________________________ CentOS-devel mailing list CentOS-devel@centos.org http://lists.centos.org/mailman/listinfo/centos-devel
Dear Karanbir.
Does anyone have details on what is going on with upstream and their HPC product ? A directory has down up on ftp.redhat.com but there is nothing inside there, and there seems to be not much going on with their sales website either.
What we could start working on in the mean time is to get a SIG together, and formulating intent / direction / mechanism etc.
Also, HPC itself is quite a vast subject - and perhaps we should not limit ourselves only to the product base that is being released upstream. So if we were to propose a HPC SIG, are there a few people here who would like to join in the effort and help ?
I would also like to join the SIG. I have a meeting at RH in late November concerning the upcoming HPC strategy and I will let you know what's going on as soon as possible.
Best Regards Marcus
On Thu, 30 Oct 2008, Karanbir Singh wrote:
Also, HPC itself is quite a vast subject - and perhaps we should not limit ourselves only to the product base that is being released upstream.
That depends on the amount of time and people interested to participate. As you say, HPC is quite a vast subject, and if only a handful of people join in I see no point in trying to achieve too much and fail.
So if we were to propose a HPC SIG, are there a few people here who would like to join in the effort and help ?
Sign me in.