Manage updates with the Download App
April 8, 2007 6:00 AM PDT
Microsoft gains technical-computing toehold
- Related Stories
-
SGI plugs in Windows for compute clusters
January 10, 2007 -
Microsoft wraps up code for 'supercomputer' Windows
June 9, 2006 -
Microsoft creating Windows for supercomputers
May 24, 2004
Windows Compute Cluster Server (CCS) runs on a group of interconnected computers that collectively tackle calculation chores. These high-performance computing clusters have swept the list of the top 500 supercomputers--but they typically run Linux, not Windows.
But when Microsoft released Windows CCS less than a year ago, the company tried to find a new niche in the market rather than go up against Linux directly. The software giant is trying to win over customers with small clusters, often integrated with the work customers are doing on their Windows PCs.
Gartner analyst
"We think that's fertile ground that nobody else has hoed yet," said Gartner analyst John Enck. "We were pretty skeptical when they came to market with this, but they're doing much better than we anticipated."
Microsoft has had some successes moving from a market in which it's strong into an adjacent market where it's not. For example, Microsoft moved from operating system software to desktop software, and from Windows on PCs to Windows on servers.
Lateral move
That's exactly what happened in the case of the South Florida Water Management District, which is using Windows CCS to power a modest-size five-server cluster that computes water flow to as part of a multibillion-dollar habitat restoration project in the Everglades National Park. The group also has a much larger Linux cluster, but the group also had Windows-based modeling tools that they moved easily to the cluster, said Akin Owosina, program manager for the district's Interagency Modeling Center.
Another reason the Windows cluster is appropriate is because outside stakeholders--everyone from the federal Fish and Wildlife Department to environmental activists--want to check model results and in some cases run those models themselves to verify the results, Owosina said.
"We want to be used by as many stakeholders and customers as we can. For many of them, the environment is Windows," Owosina said.
And for Saifur Rahman, professor of electrical and computer engineering at the Virginia Polytechnic Institute, using Linux would have required new expertise. His organization is running Windows CCS on a 16-server cluster for research in transportation and in cancer-related molecular modeling.
"We wanted to remain within the Windows environment so that we could use our existing applications and did not have to retrain our graduate students who have been working in this environment for several years with data from end users," Rahman said.
In particular, his students use the Matlab mathematical calculation and data-processing software on Windows. Matlab on the desktop can tap into Matlab on the cluster for heavy lifting.
Microsoft gives itself high marks for its results so far. "We acknowledge we have more work to do here, but we've made good progress in the first year," said Shawn Hansen, Microsoft's director of HPC (high-performance computing) marketing. "We've been very pleased with the results and the uptake."
See more CNET content tagged:
cluster, Matlab, high-performance computing, modeling, Linux
25 comments
Join the conversation! Add your comment
squashes innovation and they always have. This sounds to me like
another paid forum post by MS....APPLE, INC. creates amazing stuff
in preexisting markets, and they always have. Get a clue.
/P
2) TIE IT TO WINDOWS!!
3) Market becomes theirs.
Over and over and over and over and over again.
You'll note that nothing in the article was about how great Windows CCS is. Rather, it's all based on the fact that:
1) People are already using Windows
2) People are already using Windows-only apps.
What a crock. Nobody can get a fair shake in the tech market anymore. Inevitibly, Microsoft clamps down on any perceived threat using the bully-club of Windows.
There are lots of companies that don't need their systems available constantly and can afford the downtime that invariably comes when running Windows.
I know individuals that run 5 - 10 server Linux clusters. And it isn't unheard of for relatively small businesses (like Lowery Digital) to have fairly large (100 server) clusters.
So, IMO, writing about a few 5-20 server clusters as a Microsoft toehold appears a bit pathetic.
In fact, what I have found during customer visits is departmental or small workgroup level clusters that only have 8,16,or 32 node clusters are much more frequent occurrence (ok, these are mostly engineering/scietific design groups - so my obsv may be biased).
For good or for bad, Microsoft has gained a tremendous amount of traction with these groups, particularly by pushing it as a low maintenance-easy to use solution. I don't think it will be long we will start seeing larger CCS clusters.
msft wouldn't take that losing spin - instead they claim their clusters are easier in some sense. this is mostly FUD, though, since there are numerous linux-based clusters which basically install with a single click. the windows world, however, doesn't really grasp how easy it can be, since SOP for windows is to have a monkey reinstall when a box has a problem. of course, there are lots of monkeys available, and pretty cheaply.
I'm a linux/clustering guy (duh) and am not too worried about CCS. mainly because it's a pre-SOA approach, but also because access to a 5-node cluster is not really interesting - it might as well be a 8 or 16-core single workstation. where clusters are interesting is when they get large and are shared by multiple groups. being able to burst to 2k processors is a major win, even if your "basal" consumption rate is 10p.
But how many 2K (or more) processor clusters are there in the world?
I mean seriously... it'll all happen the same way all of MSFT's other attempts at this niche:
* Once every couple of years, MSFT tweaks the unholy crap out of Windows, builds new 'uber clust0rz!' with it, manages to sell it to someone, and proclaims the product as top dog.
* said cluster manages to crawl into (maybe) the top 10 on credible supercomputer listings.
* IBM shows up a week later with something that blows the doors off of MSFT's best effort.
* rinse, repeat.
/P
top positions on the Top 500 list. We'd get a machine into 2nd or
3rd place on the list and a couple of months later it would be 25th
or lower. Its just the nature of the industry. When you can boost a
linpack by just tossing on a few hundred more processors this sort
of thing is inevitable.
I don't really understand what all the buzz about this product is. If someone is smart enough to be a grad student, can't he or she learn Linux? It's not that difficult. None of our grad students have had trouble with it. Most know Linux when they come in to the lab anyway. This really sounds like a good product for lackeys whose strategy for improving IT is to buy EVERYTHING from Microsoft, regardless of whether it's a good product.
We're using Rocks right now and it's working quite well for us. The installation process is very simple. I can rebuild the entire cluster with a new Rocks release in a couple of hours tops. Management is simple. It's easy to add users to the cluster and install software. It also doesn't require license keys, product activation, or have WGA. I can just imagine getting a call on my cell one night from someone whose job on the cluster failed 5 days in. Then I find out it's because WGA shut down Windows due to a validation error.