Cluster Challenge
Hey, you can actually see me in this picture, arranging marketing materials to make a backstop for the slot racers (which have a tendency to fly off the track): http://www.oreillynet.com/onlamp/blog/2007/11/sc07_cluster_challenge_day_one.html The blog entry is sort of interesting, too.
//insert technobabble
I passed out about three am. ajhinz woke me around 7:30, saying he wanted to kill our GAMESS run (which had been going for four hours already). We talked it over, looked at the documentation and the state of the system, and agreed: the run had likely done no useful work, instead sucking up all of the CPU by thrashing RAM. We killed the job, and a few minutes later our advisor, Laura, showed up. I started a long Povray run, and she took Adam back to the hotel so he could get some sleep. I held down the fort for another few hours until Sarah and Valkyrie (sans Pikachu hat, today) showed up. I got them trained in on povray, and after a little distraction on the show floor, left to go get some real sleep.
//end technobabble
But alas, it was not to be. Adam and I got woken after I'd been asleep an hour and a half (and him about five) by the urgent ringing of his cell. It seems Reno's power grid puked. All was not lost, though. By the time Adam, Greg (who was sleeping in) and I got to the convention center on the bus, Andrew had gotten the machine rebooted, set up, and had manually restarted the Povray run. It had completed, and he had moved on to the next one. The outage hardly cost us - something very few of the groups at Supercomputing can claim.
At this point, I wasn't much feeling like sleeping any more, so I chatted up some folks at Penguin Computing looking for a job, or maybe an internship for my teammate, Greg. I think it went pretty well, and I got introduced to Don Becker, which I thought was really cool. Afterwards, I went back to the hotel room long enough to get a shower and a change, and come back to the convention center.
I had thought that I would man the booth alone for a while, because Penguin was having a party and invited the whole team. I decided to take the hit and let them all have the experience of going. But it turns out that Val and Sarah had gotten really into the problems they were working on, and Greg doesn't much like parties, and that Andrew was having fun running jobs on the cluster and making animations from stills. And Adam had planned on staying around with me.
So now, six hours later, we're all hanging out together. We've run several more Povray jobs, we've started a new GAMESS run, and we've eaten some pretty decent Thai food (that Stonybrook arranged to have delivered.) We've played some frisbee in the convention center, played with our slot car racers a lot, and spent a lot of energy compiling and recompiling and re-recompiling our software. The social barriers between the teams have fallen further - even with Purdue. I'm still beat, but you know, it's good. My belly and my heart are full.
//insert technobabble
I passed out about three am. ajhinz woke me around 7:30, saying he wanted to kill our GAMESS run (which had been going for four hours already). We talked it over, looked at the documentation and the state of the system, and agreed: the run had likely done no useful work, instead sucking up all of the CPU by thrashing RAM. We killed the job, and a few minutes later our advisor, Laura, showed up. I started a long Povray run, and she took Adam back to the hotel so he could get some sleep. I held down the fort for another few hours until Sarah and Valkyrie (sans Pikachu hat, today) showed up. I got them trained in on povray, and after a little distraction on the show floor, left to go get some real sleep.
//end technobabble
But alas, it was not to be. Adam and I got woken after I'd been asleep an hour and a half (and him about five) by the urgent ringing of his cell. It seems Reno's power grid puked. All was not lost, though. By the time Adam, Greg (who was sleeping in) and I got to the convention center on the bus, Andrew had gotten the machine rebooted, set up, and had manually restarted the Povray run. It had completed, and he had moved on to the next one. The outage hardly cost us - something very few of the groups at Supercomputing can claim.
At this point, I wasn't much feeling like sleeping any more, so I chatted up some folks at Penguin Computing looking for a job, or maybe an internship for my teammate, Greg. I think it went pretty well, and I got introduced to Don Becker, which I thought was really cool. Afterwards, I went back to the hotel room long enough to get a shower and a change, and come back to the convention center.
I had thought that I would man the booth alone for a while, because Penguin was having a party and invited the whole team. I decided to take the hit and let them all have the experience of going. But it turns out that Val and Sarah had gotten really into the problems they were working on, and Greg doesn't much like parties, and that Andrew was having fun running jobs on the cluster and making animations from stills. And Adam had planned on staying around with me.
So now, six hours later, we're all hanging out together. We've run several more Povray jobs, we've started a new GAMESS run, and we've eaten some pretty decent Thai food (that Stonybrook arranged to have delivered.) We've played some frisbee in the convention center, played with our slot car racers a lot, and spent a lot of energy compiling and recompiling and re-recompiling our software. The social barriers between the teams have fallen further - even with Purdue. I'm still beat, but you know, it's good. My belly and my heart are full.