Platform Computing has taken the beta off its Infrastructure Sharing Facility, a way of greasing the adoption of Windows HPC Server 2008 among the supercomputing folk. Having seen a zillion different operating systems and architectures over the past three decades, these HPC techies like the portability and commonality of Linux …
If anyone went up to one of our HPC sys admins and suggested windows HPC server they would be laughed out of town.
If anyone suggests to me that I design my jobs to run on windows HPC server I would laugh and laugh and probably never take that person seriously again.
Good luck to Microsoft in this field, they are going to need it!!! Red Hat is just too damn good at this sort of thing.
Finances are iffy for the supercomputer centre
The configuration control of clusters is a major operational issue. Deploying a second operating system not only double that problem, but -- because Microsoft has chosen not play well with Unix/Linux to maximise it's commercial computing revenue -- requires a second set of configuration control software.
Increased cluster system administration will increase cost per TFLOP. For the supercomputer centre to be competitive the Linux TFLOP cycles will need to be billed out as if it is a single-boot cluster. Leaving the Windows TFLOP cycles to carry all the additional costs, making those cycles very expensive. You need to find a customer willing to pay those higher prices before even considering such a deployment.
What kind of amateur uses Red Hat on a cluster?
Well, I use Red Hat. Currently I have 1174 jobs running in "the cloud". Probably take around 4~5 hours each.
What do you use??
Jonh 137, what do you run on your cluster?
BSD Baby all the way
Win / Lin pretenders.
read top500 and ...linux IS HPC...rocks'roll, caos-nsa,red hat cluster suite...
Windows HPC..very funny...ROFLMAO