tenfiveoh 2004-10-16 09:36:11 | Anyone have any thoughts on the Large Hadron Collider project? It seems rather ambitious. Is it a pipe dream or will in fact be a reality in three years. Do you think having LHC designs tested on the increasingly popular BOINC platform is going to take away from, compliment, or have no impact on this poject? |
[SG]Santas little helper 2004-10-16 12:48:55 | Hi tenfiveoh, good to know that I'm not the only one interested in LHC/Cern Because of that I was at the local university to a colloquium where an involved scientist spoke about the importance of LHC for the particle physics. He also told the audience that they are right in this moment at work for the LHC project. In his case it was a special part of the ATLAS project and the related detector. Let's hope that they can begin with experiments in 2k7. Sorry for the simple wording ^^ Just bein' a physics fanatic |
Stephen Brooks 2004-10-16 13:43:18 | quote: It's kinda big and kinda round and kinda nearly bankrupted CERN quote: It will happen - maybe a year late at most. They basically slashed all other projects and R&D in order to be able to concentrate totally on LHC work. quote: I hadn't heard about that until recently, though I thought they were testing it using BOINC on their own computers rather than across the internet. Oh, hang on I found a link. What they're doing is seeing which of their manufactured dipoles can go in which slots around the ring and still allow the beam to circulate for a long time (badly manufactured ones lead to instabilities). I have no idea what impact that will have on this project - there are after all a lot of different accelerators to design. However, theirs does not use AI like this one does, nor does it specifically aim to increase the beam intensity, simply to make sure the beam keeps on going. Overall I've not found other distributed projects that combine AI techniques with a real-world problem like Muon1 does, though I've not really looked. I've found plenty like "ChessBrain" and "Golem something" that are AI but not very real-world, and others like the drug design ones which are real-world but brute force. This, however, is very unfair on the likes of UD/THINK because his algorithm, while not "AI" as such, is extremely clever and his simulation a lot more sophisticated than what Muon1 does. Ultimately it just depends what project is to your taste. And even if everyone leaves this project I'll just ask around at my lab for idle Windows machines (of which they have plenty) and use the infrastructure I've already developed to continue the work. |