Cool/Weird Stuff To Do On a Cluster? 608
Gori writes "I'm a researcher at a university. Our group mainly does Agent Based Modeling of interdisciplinary problems (think massive simulations where technology, policy, and economics meet). Recently, we managed to get a bunch of money for a High Performance Cluster to run our stuff on. The code is mostly written in Java. Our IT support people are very capable of setting up a stable cluster that will run Java perfectly. But where's the fun in that? What I'm trying to figure out are other, more far-out and interesting things to do with this machine — think 500+ Opteron cores, 2 GB RAM per core, a gigabit interconnect with some badass switches, a massive storage array, plus a bunch of UltraSPARC boxes. So at times when there's no stuff to crunch, I'd like to boot the thing up with a 'weird' system image and geek around in the name of science. Try fancy ways of building models, dynamically adding all sorts of hardware to it, etc. Have different schedulers compete for resources. Imagine a Matlab vs. Boinc vs. ProActive shootout. Maybe run plan9 on it? Most of us are not CE/CS people, but we are geeky enough. So, what would be the coolest and most far out thing you would do with this kind of hardware ?"
How about you don't? (Score:5, Insightful)
And save the environment a little bit?
Coolest? (Score:5, Insightful)
"So, what would be the coolest and most far out thing you would do with this kind of hardware?"
Instead of pissing around with stuff that may not go anywhere other than a few giggles over lunch.
Why not just rent, or lend it out to people who don't have the funding or equipment that could use this cluster for a better purpose than "playing around"?
Just saying...
Re:Coolest? (Score:5, Insightful)
Its incredible a university would let this kind of equipment go to waste. If people arent clamoring to run things then you either have a non-existent (or terrible) CS department or too much money that should have gone elsewhere.
Regardless, there are tons of grid clients out there. There's always something to run.
Re:How about you don't? (Score:3, Insightful)
> And save the environment a little bit?
That is one of the sanest postings I've come across on slashdot. So why is it marked as a troll?
Stone Soup (Score:5, Insightful)
I (Mat) work at Vanderbilt University's supercomputing center. Our university supercomputing center was originally a joint venture between the proteomics and high-energy physics departments, but they decided to make it a independent university-wide initiative to bring HPC tools to all users.
Before we founded the center, there were a lot of groups that required computing on campus, but it was highly inefficient. Their local clusters had lots of free cycles (low return on asset) that they couldn't effectively share with other users, the clusters were down quite often (grad students and postdocs are poor sysadmins, plus they should be doing actual research anyway). Several other problems related to either pooling of resources or pooling of knowledge, you get the idea...
I highly recommend setting up a batch scheduler such as Torque/Maui and opening your cluster to all researchers on campus. You'd be surprised how much demand is out there. We have all the usual math/science/engineering/biomedical users, plus users in more esoteric fields (nursing, accounting, music, psychology). You can always give your group a higher job priority if needed. It gives a higher return on asset and gets lots of goodwill on campus (and, potentially, at funding agencies). You can charge users for support, storage, etc for cost recovery, or even use it as a revenue source if your grants allow.
Having different types of users also allows cross-pollination of ideas. We have a large number of biomedical researchers who are now using a high-energy physics software (geant), biomedial people who are teaching other users how to program in R, etc. These are avenues for research/discovery that didn't exist before.
Re:Vista? (Score:4, Insightful)
Re:Coolest? (Score:3, Insightful)
...But if what you normally do on the cluster/system is "boring" then maybe you are in the wrong field of investigation.
And, if you did rent/lend it out to another group of people, you may be just as interested in what they are doing with it, and what they are doing may correlate to what you are doing.
Perhaps that third party is doing the "playing around", but if you rented it out to them for something they can afford, you are still using the system for a benefit to them, as well as your goals (extra money to put back into the system, pay for power, etc).
Re:Vista? (Score:5, Insightful)
You should maybe hide your signature when you say things like that.
Itg is widely accepted that Vista is a waste of resources, and therefore all articles about powerful computer resources are going to have Vista jokes. You can't kill off a meme on your own. If you don't like repetetive humour then perhaps you should change your moderation to mod down all funny comments.
Anyway, slightly back on topic - I think they'd want to keep it a cluster rather than degenerate into a clusterfuck, so it may be better to avoid installing Vista.
Re:Reduce Your Carbon Footprint (Score:3, Insightful)
In my experience as a sysadmin, when you have a resource, your users want it to be "up" all the time, no matter what. If it's interactive, they'll leave VNC sessions or xterms or screen sessions running on it and want them to be there when they come back. If it's noninteractive (ie: a queue/batch system), your users want to be able to submit jobs now, without waiting for the sysadmin to come in and fire it all up and make it run.
Without some serious organizational political capital, it's pretty hard to pull off powering down the compute resources. It can be done, but it's going to leave a lot of people unhappy.
Re:Imagine . . . (Score:5, Insightful)
Uh. So the endless Beowulf jokes are funny to you, but the nowhere near as long running Vista ones aren't? You perhaps should go into therapy to resolve your Vista issues. I'm intrigued as to why you consider it better than XP in any way.. apart from apparently the calendar is more comprehensive than XP for tracking changes in our date system over the last while.. but other than that. I honestly don't see the benefits. I'm not being hypocritical either - I was doing fine with 98 until games started requiring XP. And I'm writing this on OS X.
Tell the physicists about it (Score:2, Insightful)
Well, you could always make a virus farm [xkcd.com]
But seriously though, let your physics department know that you have idle time on such a cluster, they will come begging at you for some cicles. I've tried running a simulation of some quantum systems from first principles, using Monte Carlo methods (my software), and it would never get anywhere on small clusters. On a big one I managed to see something, but after days of processing.
Nothing like hypocrisy! (Score:4, Insightful)
Couldn't you be doing something yourself to save the environment right now - something like turning off YOUR computer instead of browsing slashdot? Oh, wait - that would require YOU to do something. That kind of environmental activism is never as much fun as simply preaching to other people what THEY should be doing. Hypocrisy is SO much fun - carry on!
Re:Convert code from Java to something else ... (Score:3, Insightful)
Bullshit. No way, no how is C++ ten times faster than Java. You're smoking crack.
Re:Coolest? (Score:3, Insightful)
that could use this cluster for a better purpose than "playing around"?
Guys, a lot of useful pure research is mostly just playing around. Take a few ideas as a baseline to get you started, then play around until you reach that "hey, that's funny..." moment. Challenge the limits as a goal in itself, then see how things act on the edge, apply rules and discipline to your game when it gets interesting.
The only real difference between play and pure research is whether or not you keep a decent lab journal.
Re:Convert code from Java to something else ... (Score:2, Insightful)
Huh? This kind of clueless drivel marked as "insightful"???
It is doubtful that there is significant amount of overhead. Fortran could be faster for specialized math tasks, for few select cases maybe an order of magnitude faster. C++ would just crash more often, without guaranteeing ANY speedup. Competently written java code, for agent-based simulations, should be up there with compiled languages.
Re:How about you don't? (Score:5, Insightful)
because it won't make one iota of difference to the environment whether that thing is on or off, your one of those token bullshit symbolism over substance wanks. That's as stupid as advocating abolishing drag racing to save gasoline.
Re:better yet (Score:3, Insightful)
Hacking reality? That sound like Engineering to me.
Re:better yet (Score:5, Insightful)
I think reverse-engineering reality is called "Physics".
Re:Super photogenesis (Score:3, Insightful)
Re:high performance java? (Score:3, Insightful)
C.
Also, 20% is no small number. (Damn filter. Ruined my one-character post.)
Re:Coolest? (Score:2, Insightful)
FWIW, that's at least as much SGI's fault for not guiding them into buying maintenance.
For big iron systems, the sales team is typically expected to handhold the customer through the process.
So, unless your uni was Urbana Champaign (NCSA) or some other place that's been buying big iron for decades, SGI was at least as much at fault for letting the department dig itself into a hole like that.
Re:How about a game? (Score:2, Insightful)
Re:high performance java? (Score:1, Insightful)
Why, what's wrong with using Java? Really I'm not being sarcastic I want to know your reasons.
http://en.wikipedia.org/wiki/Comparison_of_Java_and_C%2B%2B#Performance
Re:Coolest? (Score:5, Insightful)
Let me see if I can help explain...
Money comes in to education through grants. Grants are usually awarded to whoever can make the best case. You'll notice I didn't say "deserving" or "beneficial to mankind" case. The best case is often the one that's written by someone who knows how to game the system be it through politics or releasing sensationalist research that may not prove anything much to anyone.
Publishing articles are kinda nice. Publishing books is usefull for a little extra income from screwing your students who have to buy the latest version you update each year. Having the media pick up your research because you just claimed women are smarter than men and the secret to cold fusion powered cars that also run on water is far, far better.
Once you have this money, it's yours to do with as you wish. Just one thing. Don't ever, ever let it be seen that you didn't really need it. If that happens, how much budget do you think the university is going to give you on the years where your grant applications fail? How about those you beg for money from, next time? If you got money to buy a supercomputer, you need it, 24/7, until you declare it's obsolete. If you don't, they might ask why a pretty-nifty-computer wouldn't have worked just as well. And so, for that reason, no one else gets time on it. Hell, you might only need it because it turns out it's the perfect sized doorstop for a door you had... but let anyone think you don't need it and you're screwed in the magical world of academia.
It's in conflict with the notion of academia being "for the good of mankind" but no more so than the notion of government being "for the people." Both simply serve the people within it... researchers or politicians. The rest is set dressing to ensure others keep paying for it.
Re:How about you don't? (Score:4, Insightful)
figure out a way to run 5000 simultaneous desktop environments on those 500 processors... and that might help the environment quite a lot
Re:Super photogenesis (Score:4, Insightful)
As well as all porn being thumbnailed, why not find the tiles for it in HD resolution instead? Or better yet, in 10000X10000 pixels?
So in essence, such a program would spit out every frame of every movie, in better than current HD resolution, that will ever be made by humans (or aliens, or anything). It would also contain the entire set of human books, past and future, and alien books for that matter, along with the correct method to translate it. It would have every bit of knowledge possible to fit into a 2 dimensional representation written down for us to read right now.
Which is all very exciting until you realise that it is, of course, not possible to do with computers in any meaningful timescale within a Universe, and for every correct set of blueprints for a time machine there must be an almost infinite set of incorrect ones which are indistinguishable from all the others until you build it.
As an interesting aside, I wonder what size bitmap it IS capable for - 2X2? 4X4?
Re:How about you don't? (Score:3, Insightful)
I respectfully suggest that 'saving the environment' while laudable, needs to be taken in context. Sure, don't drive to the mall with your friends, all in your own cars. Do, however, keep doing things to advance the state of our knowledge - geeky fun with a massive load of hardware, for example, unless we are talking 'substantial' environmental impact. What constitutes 'substantial' is, of course, subjective, but at least think about it and don't just say, "Well, it uses energy and gives off heat - BAD!"
To those who would snark - what are YOU doing RIGHT NOW to save the environment? Reading /.! Hypocrite!
Re:Super photogenesis (Score:2, Insightful)
Re:How about you don't? (Score:4, Insightful)
"because it won't make one iota of difference to the environment whether that thing is on or off..."
assuming it's 500 cores, and also assuming (conservatively) that the diffrence between idle and full load is only 100 watts per processing unit - that means about 50 extra KWh consumed by this thing at near full utilization.
X 24 hours = about 1.2 million watts.
1 short ton of coal yields about 2500 KWh of electricity at average efficiencies.
If I've done the math right, you can imagine dumping an extra 1/2 ton of coal on a fire somewhere to run this thing (at load) for 1 day.
According to DOE - Burning coal produces 2.117 lbs of carbon per KWh. So even 1 hour at full load introduces an additional 50 pounds of CO2 into the atmosphere
Again, all assuming this cluster sits somewhere (like america) where most of the electricity gets generated from coal or other fossil fuels. YMMV.
Important to remember - there isn't any storage or margin in the power grid. Every time turn on a light switch or run a CPU up to max with SuperPrime, somewhere a turbine starts turning that little bit faster - it's always got to be nearly in balance.