Tweet
Part 2 in the "whine about Chinese academics" series (continued from DaShi's thread)
The department here hosts a high-performance computing cluster containing a thousand processors or so. Everyone in the department, as well as a few researchers at collaborating institutions have basically unrestricted access to it (you can submit as many jobs on as many processors as you want, on a first-come first-served basis). I submitted a couple of small jobs running on like 5 processors at around noon (runtime ~1 day) and had another few (~10 processors' worth) ready to go when I got a mass email from a full professor who requested that all users hold off on submitting any more jobs for a few hours because he had some time-critical processes to run before he presented the results at a conference next week.
So, being the reasonable fellow I am, I didn't submit my extra jobs (and even considered killing the ones I'd already sent). I can wait a few extra hours to get my results, but it does inconvenience me a bit. I check back a couple of hours later, and 3 different Chinese grad students (who don't work for the professor in question) have submitted enough jobs to choke up the entire cluster for a day or so.
Holy ****, *******s. These kids are like 2nd or 3rd year grad students. Do they not realize how ****ing stupid it is to piss off a FULL PROFESSOR, or how ****ing low on the totem pole they are? The prof could easily request that these kids have their access to the cluster restricted.
The department here hosts a high-performance computing cluster containing a thousand processors or so. Everyone in the department, as well as a few researchers at collaborating institutions have basically unrestricted access to it (you can submit as many jobs on as many processors as you want, on a first-come first-served basis). I submitted a couple of small jobs running on like 5 processors at around noon (runtime ~1 day) and had another few (~10 processors' worth) ready to go when I got a mass email from a full professor who requested that all users hold off on submitting any more jobs for a few hours because he had some time-critical processes to run before he presented the results at a conference next week.
So, being the reasonable fellow I am, I didn't submit my extra jobs (and even considered killing the ones I'd already sent). I can wait a few extra hours to get my results, but it does inconvenience me a bit. I check back a couple of hours later, and 3 different Chinese grad students (who don't work for the professor in question) have submitted enough jobs to choke up the entire cluster for a day or so.
Holy ****, *******s. These kids are like 2nd or 3rd year grad students. Do they not realize how ****ing stupid it is to piss off a FULL PROFESSOR, or how ****ing low on the totem pole they are? The prof could easily request that these kids have their access to the cluster restricted.
Comment