Slashdot Mirror
Sun Donation Spurs Linux Cluster at Purdue
An anonymous reader writes "Purdue University, with a $3.6 million gift from Sun Microsystems, is giving recycled PCs new life as a computer cluster that makes high-performance computing power available in undergraduate classes. 'Previously, my students could only do what I'd describe as 'proof' animations - small, low-resolution and not presentation quality,' [Professor Richard] Paul said. 'With access to this computing power, the students will be able to ship their software files of instructions to the Linux cluster, and it will come back in three or four hours with modeling, lighting and animation. Students will get to experience the whole thing in terms of scale and presence, and they can do longer animations.' More images of the current Linux cluster and other servers at Purdue are out there."
Perhaps person was referring to files saved by various animation programs, not necessarily actual program code. Not accurate in either case, but what's the big deal?
You gonna leave your long-running render on a lab computer unattended?
Blar.
Why donate computers to a university, as opposed to the poor in other continents or even in own country? Just so they could have a bigger cluster for animations? In my eyes, that doesn't make sense.
What's in it for Sun?
Sun donated the 6800s and E10k. The 6800s are one of the first sites where Sun has installed the SunFire Link interconnect, high-bandwidth, low-latency product for clustering. The addition of these resources freed up the PCs so they could be used for Linux cluster. At one point Sun wanted to donate V60x's but Purdue didn't have the room and didn't want them so they could reuse the PCs.
The high performance computing cluster will be used for many things.... to show both the value of large SMP systems as well as the value in a beowulf-style cluster (most likely running Sun Gridware) for computing which can be distributed.
Sun and Purdue have always been very close and this deal helps to renew their relationship. There are stories of a Sun 3/60 (or 4/110?) being strapped to the back of a tractor for the agricultural school being used for seed, pestitcide and herbicide mixture and distribution.
Sun appears to be positioning themselves lately to take on the Linux market. If they really cut a deal with fijitsu to move more of their Solaris work to Japan, maybe, just maybe, Sun is trying to move towards Linux. That would help explain why Bill joy left.
As to lack of money, no. This company is a long ways away from being Corel. They have money, just not to waste.
I prefer the "u" in honour as it seems to be missing these days.
Redundant, I know, but it has to be said
Why do so many people think this has to be said?
In case you dont know, grants generally specify how the money is to be spent, and, as in cases like this, what is going to be researched. As I said, nothing new, nothing out of the ordinary.
The second sounds like an ordinary business deal. Since MS is out to make money just like every other business, why should they ignore the bottom line? If they *had* donated the resources AS WAS REQUESTED, you would be complaining that MS was giving away stuff in order to replace Linux.
Sounds like yet another of those 'cant win' situations you guys keep writing up for MS. Post- .bomb companies need to make money, in case you hadnt heard.
Manipulate the moderator system! Mod someone as "overrated" today.
Just a curious economics question (my MBA is acting up). Is there any market for those cycles outside the University? If so, would it not have made more sense financially to sell the extra cycles to company X and use the extra cash to buy a full time computer(s) for the research that needs extra cycles? I'm sure United Devices didn't work for free so someone is paying for that (research funds?). How much time do your researchers have to spend getting the software apps to work in a distributed ("grid") mode, which takes away time/money from basic research. As someone who teams with a certain large University to do Gov't software research we often see money in the University budget for special hardware or software. We don't always approve it, but if it is well justified it goes thru. If at all possible I want my researchers to have what they need, not be sharing unpredictable spare cycles. Of course some is better than none, and I don't know how you allocate your costs of the spare cycles to research. It may be that having their own systems is cheaper in the long run. I'm assuming someone did the Math on that and also did a prototype of the grid model to see if it works for the researchers at Purdue. Any info you can give as to the cost/benefits and how it works out would be appreciated, as if it works there I may ask some of my researchers to use it!