For Windows I couldn’t find anything.
If you google “Windows supercomputer”, you just get lots of results about Microsoft supercomputers, which of course all run on Linux.
Microsoft earnestly tried to enter the space with a deployment system, a job scheduler and an MPI implementation. Licenses were quite cheap and they were pushing hard with free consulting and support, but it did not stick.
Yeah. It was bad. The job of a Supercomputer is to be really fast and really parallel. Windows for Supercomputing was… not.
I honestly thought it might make it, considering the engineering talent that Microsoft had.
But I think time proves that Unix and Linux just had an insurmountable head start. Windows, to the best of my knowledge, never came close to closing the gap.
Windows is a per-user GUI… supercomputing is all about crunching numbers, isn’t it?
I can understand M$ trying to get into this market and I know Windows server can be used to run stuff, but again, you don’t need a GUI on each node a supercomputer they’d be better off with DOS…?
Oh yes! To be clear - trying to put any version of Windows on a super-computer is every bit as insane as you might imagine. By what I heard in the rumor mill, it went every bit as badly as anyone might have guessed.
But I like to root for an underdog, and it was neat to hear about Microsoft engineers trying to take the Windows kernel somewhere it had no rational excuse to run, perhaps by sheer force of will and hard work.
There was a time when a bunch of organisations made their own supercomputers by just clustering a lot of regular computers:
https://en.wikipedia.org/wiki/System_X_(supercomputer)
For Windows I couldn’t find anything.
If you google “Windows supercomputer”, you just get lots of results about Microsoft supercomputers, which of course all run on Linux.
No there was HPC sku of Windows 2003 and 2008 : https://en.m.wikipedia.org/wiki/Windows_Server_2003#Windows_Compute_Cluster_Server
Microsoft earnestly tried to enter the space with a deployment system, a job scheduler and an MPI implementation. Licenses were quite cheap and they were pushing hard with free consulting and support, but it did not stick.
Yeah. It was bad. The job of a Supercomputer is to be really fast and really parallel. Windows for Supercomputing was… not.
I honestly thought it might make it, considering the engineering talent that Microsoft had.
But I think time proves that Unix and Linux just had an insurmountable head start. Windows, to the best of my knowledge, never came close to closing the gap.
But, surely Windows is the wrong OS?
Windows is a per-user GUI… supercomputing is all about crunching numbers, isn’t it?
I can understand M$ trying to get into this market and I know Windows server can be used to run stuff, but again, you don’t need a GUI on each node a supercomputer they’d be better off with DOS…?
I could see the NT kernel being okay in isolation, but the rest of Windows coming along for the ride puts the kibosh on that idea.
Oh yes! To be clear - trying to put any version of Windows on a super-computer is every bit as insane as you might imagine. By what I heard in the rumor mill, it went every bit as badly as anyone might have guessed.
But I like to root for an underdog, and it was neat to hear about Microsoft engineers trying to take the Windows kernel somewhere it had no rational excuse to run, perhaps by sheer force of will and hard work.
Yeh it was system x I worked on out default was redhat. I forget the other options but win and mac sure as shut wasn’t on the list