Posts by proxima

21) Message boards : Number crunching : Laptop cleanup doubled output (Message 39113) Posted 7 Apr 2007 by proxima Post: Thought I'd tell this story in case it helps anyone else. Sorry if it's stating the obvious to some people, but someone might appreciate the tip. I recently noticed that one of my laptops, a Dell Inspiron 1150, has a very low RAC compared to what I'd expect. Looking closer, I noticed that the CPU benchmark shown in BOINC is only around 600 most of the time, but when switched on from cold, can be as high as 1350 before reducing to 600 again soon after. Using "ThrottleWatch", I discovered that the CPU is throttled back to 50% most of the time, and sometimes down as low as 25%, reducing a 2.6GHz machine to a 650MHz one at times. No wonder its output has been low. Using "hmonitor", I confirmed what I thought: The CPU (Intel Celeron) was running at up to 75C, and having to throttle back to get below 72C. I took the keyboard off and removed the heatsink assembly, cleaned a lot of dust out of the heatsink, cleaned up the old dried-on thermal paste, and reapplied some Arctic Silver 5. Put the whole lot back on, and the CPU has been running at 100% ever since, max temperature down from 75C to 58C. Not surprisingly, the fan speed is much lower, so the whole thing is quieter. Lesson: If you have a machine running slower than you think it should, use software like ThrottleWatch and hmonitor to see if it's too hot and throttled back. If so, and you're comfortable working with hardware, then a good cleanup of the heatsink and new thermal paste may make a big difference. In my case, it's doubled the output of one of my machines.
22) Message boards : Rosetta@home Science : MS-Disease (Multiple Sclerosis) (Message 37634) Posted 9 Mar 2007 by proxima Post: I don't remember the statements that came up on previous MS requests I found this statement at Wikipedia : With regard to the message board posts, we aren't yet doing any work on diabetes or Multiple Sclerosis specifically, but if we can generate accurate structures of proteins involved in these diseases using the methods you are helping us to develop, it will contribute to efforts to develop therapies. The way I understand it, we're not working on MS specifically (I wish we were, by the way), but if Rosetta@Home is successful, the fight against disease in general will be revolutionised.
23) Message boards : Number crunching : Predictor of the day (Message 37465) Posted 5 Mar 2007 by proxima Post: Yes, thanks for getting it going again. I guess I'm not the only geek who checks fairly often to see if I'm the POTD or not...
24) Message boards : Number crunching : there was work but your computer doesn't have enought memory (Message 37381) Posted 4 Mar 2007 by proxima Post: I'm getting the same error, having not changed any settings or the BOINC version. I'll follow your suggestion and change both memory limits to 95% or something - also, I understand Matt's suggestion that the admins need to release some jobs that run in 256MB. Until then, 3 of my 6 machines are idle...
25) Message boards : Number crunching : Predictor of the day (Message 37322) Posted 2 Mar 2007 by proxima Post: The 'predictor of the day' seems to have stopped since Feb 13th - will this be restarted at some point? I think it's a useful incentive to some crunchers (myself included...)
26) Message boards : Number crunching : Predictor of the day! (Message 35481) Posted 25 Jan 2007 by proxima Post: A related question: Has there ever been a "Predictor of the day" who wasn't a member of any team? From a quick glance at the archive, it looks like all 'predictor of the day's have been members of a team - am I the only person who crunches independently? Of course, the 'predictor of the day's are more likely to be big-hitters, who perhaps are also more likely to be team members than independents.
27) Message boards : Number crunching : Failed to stop applications; aborting CPU benchmarks (Message 32692) Posted 15 Dec 2006 by proxima Post: Will do - I'll see how this goes, and try the new version if necessary. Thanks. Tom
28) Message boards : Number crunching : Failed to stop applications; aborting CPU benchmarks (Message 32686) Posted 15 Dec 2006 by proxima Post: I wouldn't let an issue like not having benchmarks run successfully determine which projects I run. Oh, I agree - me neither. But when this error occurs, all processing stops - no further work units take place until I stop and restart BOINC. As my machines run unattended most of the time, it can be days before I notice, and with this happening probably at least weekly, my machines are probably running less than 50% of the time overall - not the whole idea! I'll keep an eye on the machines now I've changed that setting to "yes", and post what happens either way. Thanks for your reply. Tom
29) Message boards : Number crunching : Failed to stop applications; aborting CPU benchmarks (Message 32643) Posted 14 Dec 2006 by proxima Post: Hmm - I've just noticed, in conjunction with re-reading one of the other threads, that I'm set to "Leave applications in memory while suspended?: no". I'll try setting this to "yes", and see if that helps.
30) Message boards : Number crunching : Failed to stop applications; aborting CPU benchmarks (Message 32642) Posted 14 Dec 2006 by proxima Post: For some time, I have been finding Rosetta/BOINC stops with errors like the following: 2006-12-09 18:44:16 [rosetta@home] Started download of file 1a19A_R30_R55_cheat.bar 2006-12-09 18:44:19 [rosetta@home] Finished download of file 1a19A_R30_R55_cheat.bar 2006-12-09 18:44:19 [rosetta@home] Throughput 145 bytes/sec 2006-12-09 18:44:20 [---] Rescheduling CPU: files downloaded 2006-12-09 19:36:54 [---] Suspending computation - running CPU benchmarks 2006-12-09 19:36:54 [rosetta@home] Pausing task PSH_0097_looprlx_GP120_OD1_138_147_7082_1414_13_2 (removed from memory) 2006-12-09 19:36:54 [---] Suspending network activity - running CPU benchmarks 2006-12-09 19:36:57 [---] Running CPU benchmarks 2006-12-09 19:37:05 [---] Failed to stop applications; aborting CPU benchmarks 2006-12-09 19:37:06 [---] Resuming computation 2006-12-09 19:37:06 [---] Rescheduling CPU: Resuming computation 2006-12-09 19:37:06 [---] Resuming network activity 2006-12-09 19:37:06 [---] Process 23015 not found Nothing then happens until I manually stop and restart the processes. There are a few references to this error on the forums, but no real solution posted as far as I could understand. I'm using BOINC 5.4.9 on Fedora Core 5. Any ideas? I love Rosetta@Home, and I'm trying to avoid moving to other projects if at all possible. However, if I cannot get this to run reliably without manual intervention every few days, I may need to.
31) Message boards : Rosetta@home Science : Rosetta video/outreach project (Message 28657) Posted 29 Sep 2006 by proxima Post: Great video; that's really good. I hope you manage to get it "out there" and seen by as many people as possible.
32) Message boards : Rosetta@home Science : DISCUSSION of Rosetta@home Journal (2) (Message 27275) Posted 18 Sep 2006 by proxima Post: Can we expect a stream of similar "top participants" updates now, as the other CASP7 results come in? Seeing how close we are (or aren't) to each target is even more motivating than the stats, in my opinion (especially when we're as close as that first one looks!)
33) Message boards : Rosetta@home Science : Time for a new Rosetta Newsletter? (Message 26242) Posted 7 Sep 2006 by proxima Post: Maybe it is time for a new news letter to former members. Anders n Certainly could be a good idea - the last newsletter is what brought me back, and I've had 6 machines on the project (mostly 24x7, or as close as I can keep to that) ever since. I had originally left the project due to instability - the newsletter promised that the project had become more stable, and from my point of view, it certainly has - I've had no problems at all. And the stats are now (apparently) fairer, too. So newsletters certainly can work.
34) Message boards : Number crunching : "Predictor of the Day" question... (Message 22029) Posted 8 Aug 2006 by proxima Post: It's not strictly each and every day. It's really the best predictor for each type of work unit Rosetta is studying. So, generally they complete a few each week and announce the best predictions received at that time. Thanks for the information.
35) Message boards : Number crunching : "Predictor of the Day" question... (Message 22021) Posted 8 Aug 2006 by proxima Post: ...and now the Predictor of the Day seems to have stopped being updated altogether? Any reason for this?
36) Message boards : Rosetta@home Science : FAD science courtesy of Caroline (Message 10805) Posted 16 Feb 2006 by proxima Post: Likewise, I put the RAR file up on http://www.proxima.org.uk/Archive.rar
37) Message boards : Rosetta@home Science : FAD science courtesy of Caroline (Message 10656) Posted 11 Feb 2006 by proxima Post: Anyone willing to provide a secondary downloading site please say so ASAP. Carlos? Proxima ? (Nec is famous for accidentally deleting everything except his own posts.) Yep, I'm more than happy to store it on my web site for backup/download purposes. I'll talk to Nec on FadBeens, and if we go for that, I'll let you know where I put it... Thanks for all the work, anyone who was involved in this, anyway.
38) Message boards : Rosetta@home Science : FAD science courtesy of Caroline (Message 10533) Posted 7 Feb 2006 by proxima Post: We had a member at FAD who was kind and courteous to provide the project with info on protein targets for various diseases. After the closure of the project the posts can not be made at its forum. They have been moved to the FaDBeens forum. Please check the posts under the username jean to see this info. Great to hear that Caroline's threads have been moved, and will be continued. I wasn't aware that had been done, and was talking to Keith about whether they could be moved! Good news, anyway.
39) Message boards : Rosetta@home Science : Well, you said you wanted feedback ... (Message 6849) Posted 20 Dec 2005 by proxima Post: Sorry to join this discussion half-way through - I only started crunching here yesterday, having left Find-a-Drug on Friday. The talk of randon-number generators is interesting, as a project of my own very nearly failed due to a bad Random Number Generator (although it wasn't RAN3 or similar - it was the 'noddy' RNG built into the C compiler I was using - that was my mistake). My application was a Genetic Algorithm, which makes very heavy use of the RNG in precise patterns (e.g. pulling off 1000 random numbers, then 3 more, then repeating, 1000, 3, etc). I was getting very strange results, which indicated only some numbers were ever getting chosen in some parts of the sequence - sometimes the "3 more" numbers were absolutely ALWAYS odd numbers, no matter how long I let it run, but the next time, they would always be even, or always divisible by 5 - weird things like that. Because of how the GA worked, this broke things quite badly. I spent ages looking for coding errors, but in the end, someone suggested I change my RNG. I downloaded an implementation of the Mersenne Twister, which is fast, has an extremely long period, high 'dimensionality', and other desirable properties that I only half-understand (not being any kind of maths expert). Immediately, all such problems were solved and I never saw any other such skewing of numbers again. I don't know how the Mersenne Twister compares to RAN3, though.
40) Message boards : Cafe Rosetta : Find-a-Drug Refugees (Message 6805) Posted 19 Dec 2005 by proxima Post: Another Find-a-Drug refugee, signing up... At least, I think all my machines are running Rosetta@Home properly - will leave it for now and see what's happening tomorrow. Cheers, Tom (proxima)

Previous 20