Posts by J Langley

21) Message boards : Number crunching : BOINC 5.10.x?? (Message 20029) Posted 8 Sep 2008 by J Langley Post: 6.2 is the latest stable BOINC release from Berkley. Howevere, LHC@home has a problem with people using these clients. If you want to use 6.2, make sure you\'re not asking for work from LHC@home. If you want to crunch for LHC, for now, you must be on 5.10
22) Message boards : Number crunching : computing on CUDA with Nvidia (Message 19962) Posted 1 Sep 2008 by J Langley Post: Hi, AFAIK sixtrack is FORTRAN-code and the project-team was lucky to find a compiler, which produced identical results on windows && linux. So we have no PPC application and I doubt we will see a CUDA (or Cell) application (for sixtrack). But that is just the information I gathered from reading the message board. Michael Also, LHC@Home has far more computer power attached than it needs, so the need for CUDA is less than at, say, Rosetta.
23) Message boards : Number crunching : The new look bugs (Message 19866) Posted 4 Aug 2008 by J Langley Post: The apostrophe and quote issue has also affected Einstein@Home, but they seem to have resolved it: http://einstein.phys.uwm.edu/forum_thread.php?id=6790
24) Message boards : Number crunching : The new look bugs (Message 19848) Posted 29 Jul 2008 by J Langley Post: It appears the server code cannot handle apostrophes in posts, it replaces them with backslash apostrophe. Presumably a botched attempt to prevent SQL Injections is causing this.
25) Message boards : Number crunching : Please note: this project rarely has work (Message 19847) Posted 29 Jul 2008 by J Langley Post: Windows or Linux doesn't make any difference as to whether or not you will get any work. It might make a difference to crunching times - I can't remember how much variability the OS causes here.
26) Message boards : Number crunching : The new look bugs (Message 19804) Posted 14 Jul 2008 by J Langley Post: Since the server upgrade, the first column of the forum (that indicates whether the thread is sticky) does not size properly in IE7 (it is fine in Opera 9.51). In IE7 and Opera 9.51 the input text box when posting does not correctly align (it spans 1 column instead of 2) and it doesn't properly wrap text (I had to copy and paste this from Notepad to see what I was typing!).
27) Message boards : Number crunching : How often does LHC shut down? (Message 19627) Posted 16 May 2008 by J Langley Post: The work we do here is valuable to the LHC... But apparently not valuable enough for them to actually pay for it. Do LHC/CERN seriously expect that in return for a base URL, you will provide them with a free supercomputer? If that's the case, I'm seriously considering quitting LHC@home in favour of projects where the scientists demonstrably appreciate the work of their IT staff and BOINC contributors. A further thought - why do you and Alex bother with LHC@home? Doesn't QMUL have some other projects it could use our computers for? (Or is the GridPP donation restricted to only running LHC@home?)
28) Message boards : Number crunching : Very short Intel computation vs. normal AMD (Message 19615) Posted 14 May 2008 by J Langley Post: I ran both machines for a couple of years before my dissatisfaction over other things finally prompted me to mothball the AMD machine. To this day, I won't even consider an AMD processor-based machine. Memory doesn't serve me well enough to be able to quote exact clock speeds and whatnot, but I do remember that I came to the same conclusion as others on this thread... that the mathmatic units on the two supposedly-comparable chips were wildly out-of-sync, with the Intel chip being the better of the two, by far. Cheers Bob in Boise On some projects, the L2 cache size plays a very big factor in WU runtimes, even when FLOPs in benchmark test are similar. This could be another example of that.
29) Message boards : Number crunching : needs 13 GB? (Message 19614) Posted 14 May 2008 by J Langley Post: I thought I'd check in case I'm being snowed... Is this really necessary? 12/05/2008 11:02:51 PM\|lhcathome\|Message from server: No work sent 12/05/2008 11:02:51 PM\|lhcathome\|Message from server: (there was work but you don't have enough disk space allocated) 12/05/2008 11:02:51 PM\|lhcathome\|Message from server: No disk space (YOU must free 13335.5 MB before BOINC gets space). Review preferences for minimum disk free space allowed. I think this message is a result of general settings. BOINC will leave a certain % of disk space free on your harddrive (and only use a certain amount in total across all BOINC projects), but if the LHC WU were downloaded the harddrive would be too full according to one of these rules. To resolve this, you can adjust the %s, or delete some BOINC data from other projects, or delete some non-BOINC data.
30) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19561) Posted 28 Apr 2008 by J Langley Post: arrgh... that could be 2.5 - 3days or more; what is the report deadline, it'd need to be nearly a week to be safe? I got another Muon WU last night, the deadline was 1 week.
31) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19556) Posted 27 Apr 2008 by J Langley Post: arrgh... that could be 2.5 - 3days or more; what is the report deadline, it'd need to be nearly a week to be safe? Deadlines are very tight on slow / non-24x7 machines (can't remember the exact length) - my 20 hour WU forced BOINC into EDF almost straight away. On the other hand, BOINC has managed the resource sharing correctly, so I haven't downloaded another WU again since.
32) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19543) Posted 25 Apr 2008 by J Langley Post: has anyone worked out the minimum system specification for running those Muon tasks? that is minimum memory, cpu and operating system. CPU varies quite a lot. My Core2 T5470 1.6GHz has completed some Muon WUs in about 1 hour, but the last WU took just over 20 hours. (The progress bar doesn't track well either, apparently a limitation of the wrapper.)
33) Message boards : Number crunching : How often does LHC shut down? (Message 19538) Posted 25 Apr 2008 by J Langley Post: Sadly a) that's not how it works and b) LHC doesn't actually have much money CERN != LHC ;-) If CERN doesn't fund LHC, who does? QMC? If CERN doesn't fund LHC, how come they get to tell you to take the forums down for 10 days (which messed up several thousand BOINC installs)?
34) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19532) Posted 24 Apr 2008 by J Langley Post: Paul, Orbit now has work for Intel Macs: http://orbit.psi.edu/oah/forum_thread.php?id=87
35) Message boards : Number crunching : How often does LHC shut down? (Message 19527) Posted 23 Apr 2008 by J Langley Post: We hope to have some summer students this year too... Doing what? The collider should be up and running by then, so surely SixTrack will be redundant; and if there are no new apps in the pipeline, won't the project close?
36) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19416) Posted 14 Apr 2008 by J Langley Post: Orbit recently relaunched. It's still in alpha, so WUs are few and far between, and mainly for Linux at the moment... Well, that leaves me out twice ... :) Only two OS-X machines and two WIndows machines ...[/quote] According to their homepage, Windows WUs will be launched this week (about 2 weeks after Linux ones), with OSX to follow (no time period specified). So it might be worth keep an eye on this project.
37) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19393) Posted 13 Apr 2008 by J Langley Post: I really did like the idea of Orbit@Home, shame it never got off the ground ... I mean, the site was launched before I left two years ago ... Orbit recently relaunched. It's still in alpha, so WUs are few and far between, and mainly for Linux at the moment...
38) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19378) Posted 12 Apr 2008 by J Langley Post: I prefer "physics" type research (given my druthers) and this and EAH are the only two fully oriented in that direction. Spinhenge and orbit@home are two more physics projects, but they are still Beta and Alpha respectively. The standalone (i.e. non-BOINC) Muon app is Production level I believe, it's the BOINC wrapping by yoyo that is Beta.
39) Message boards : Number crunching : project encountered internal error (Message 19364) Posted 11 Apr 2008 by J Langley Post: Again, the fundamental problem, beat to death elsewhere, is that the projects, in essence, treat the community (for all the praise the heap on us in their advertising videos {here I am thinking specifically of the one done by an LHC guy at a conference and the RAH one on thier site}) we are treated as just worker bees, anonanymous, interchangeable, and ultimately, disposable... There was a burst of activity across projects when some participants (e.g. Akos F, and Crunch3r) produced optimised apps for projects, but this definitely seems to have died down of late. extrodinary loss rate of new participants. I find it a little depressing that so much of BOINC's time is devoted to S@H myself. If you're looking for more personal involvement, perhaps projects such as Stardust@home and GalaxyZoo would be more to your taste?
40) Message boards : Number crunching : Muon accelerator design via BOINC (Message 19363) Posted 11 Apr 2008 by J Langley Post: Yes, and I just dropped Rosetta@Home because I cannot figure out if it is Beta or not, and the postings don't explain much ... And it appears they "slipstreamed" in a new version and though it did not hit me ... it did cause problems for others. I have enough troubles ... So, I shifted priorties again ... And, LHC is out of work again, and I cannot report my last 10 tasks ... sigh ... As I understand, Rosetta is Beta in the sense that the proteins which are folded there are processed so that different protein folding algorithms can be tested. Once a good algorithm has been found, then that algorithm will be used for biological / medical science. WCG and F@H seem to be doing real science now. SIMAP is also doing real work, but only for a week or so each month, and their project is one that doesn't lend itself well to community involvement (beyond the donation of CPU cycles) now that their applications have been optimised. The bulk of my BOINC time dedicated to bio-projects is now split between SIMAP and POEM@home (who are working on the same stuff as Rosetta, but seem to be a bit more involved with their users, though this could be just the first flush of interaction that most projects manage).

Previous 20 · Next 20

LHC@home