Google and IBM to Provide Cloud Computing to Students

John "butter/oreo" Bajana-Bacall writes to tell us that IBM and Google have decided to team up to provide cloud computing resources to participating college students. "Most of the innovation in cloud computing has been led by corporations, but industry executives and computer scientists say a shortage of skills and talent could limit future growth. 'We in academia and the government labs have not kept up with the times,' said Randal E. Bryant, dean of the computer science school at Carnegie Mellon University. 'Universities really need to get on board.' Six universities will be involved in the initiative. They are Carnegie Mellon, Massachusetts Institute of Technology, Stanford University, the University of California, Berkeley, the University of Maryland and the University of Washington."

Comments

Re:What if & modern computing

[Anonymous Coward]

But to those of us fortunate to be "Computer Wizards" who live and work at the Computer Utility, life would be grand!

And here you have identified what I believe is the root cause of the Microsoft hatred around here. The "Computer Utility" is not really any different than the "Glass Computer Rooms" of the 70s and early 80s before the PC came along. So now many of those "Computer Wizards" who formerly inhabited the glass rooms are dyed-in-the-wool Microsoft bashers largely because of Microsoft's role in providing computing power "for the rest of us" (to borrow an Apple phrase).

Meta question

[mosel-saar-ruwer]

Presumably these clusters are for really hard problems - folding proteins, or simulating nuke explosions, or searching for exotic primes, or classifying Lie Groups, or proving Four Color theorems, or whatever - i.e. presumably these programs are expected to run for a long, long time before they terminate.

On the other hand, a fellow named Alan Turing once proved that we can't know whether an arbitrary program will ever terminate.

Now here's the question: If you allow a student onto one of these clusters, and if his program keeps running and running and running and running, with no apparent end in sight, then how do you know whether there's actually an infinite loop within his program, or whether it's just a very, very, very hard problem he's trying to model?

So if you are one of the lucky few who gets chosen [or at least pre-selected] for this sort of thing, then will you have to submit a "proof" of the finiteness of your program before you're given the green light?

And will they provide any formal "template" within which the student could "prove" finiteness, or at least offer an outline of [a hope for] a proof?

And might there be some set of "mileposts" which the program is required to meet in a given time, and if, as it runs, the program fails to meet a milepost in time, then it's given the heave ho?

In a similar vein, are the lucky few required to "prove" that they have used all of the fastest known algorithms for each of their calculations?

Just as an example, have you ever timed the computation of "n choose k" using the actual multiplication & division of the factorials, and then compared it to the speed of Pascal's triangle?

Or tried anything in signal analysis without the benefit of O(nlog(n)) algorithms?

Re:Cloud computing?

[jlarocco]

Does that really need a new buzzword? Sounds like the same old shit that people have been doing with the internet for 10 years now. At the very least, isn't that basically the definition of "Web 2.0"? What's the difference?

I was a little worried I had completely missed out on some new phenomenon, but that Wikipedia page has only been around since March. Sounds to me like Google and IBM just want to inspire "OMG!!1 We're missing out on 'cloud computing'!1" in idiot PHBs and investors.