Posts by Tom Philippart

21) Message boards : Rosetta@home Science : CASP 8 (Message 52873) Posted 5 May 2008 by Tom Philippart Post: The last CASP this project went through, there was much more fanfare and many more announcements leading up to the event. How come I don't see this now? It's kind of sad, since we had the huge member increase which more than doubled the project Tflops, the project became much more silent. Dr Baker still has his blog, but the project team doesn't reply on the forums that much anymore. They even systematically ignore some questions, like CPU optimization/Xbox development/specific disease research. It really depressed me because I see/saw a huge potential in this project. I'd be glad if they finally restart talking actively to their contributors. That's a reason why some people are severely holding back their crunching power to support other projects (like POEM, which also participates at CASP. (but nobody seems to mind this bad public image that has begun to spread) CASP7 was an awesome crunching time, everybody loved it and many teams suspended other projects in order to support rosetta. I once got an excuse for the lack of communication lately saying that the team was too budy with the user increase... Still communication is crucial in DC imo. I hope the project can recreate the fun and enthusiasm it once had through a CASP announcement in the news section and an attempt to answer the most frequent questions (noted above) sincerely
22) Message boards : Rosetta@home Science : DISCUSSION of Rosetta@home Journal (3) (Message 52417) Posted 13 Apr 2008 by Tom Philippart Post: Happy to learn that Rosetta will be updated for CASP8. Will it be the traditional code or the new Minirosetta? Wouldn't it make more sense to put all new efforts into the new code, or is it not stable enough for that yet? Either way, I'm looking forward to seeing the results of CASP 8. Yes, we are putting almost all of our efforts into the new code, and we have a lot of testing to do in the next three weeks! Any chance of including SSE/SSE2/SSE3/... optimizations?
23) Message boards : Number crunching : Multiple 5.81 - BOINC_SYMM_FOLD_AND_DOCK_RELAX - crystal_foldanddock errors (Message 49066) Posted 26 Nov 2007 by Tom Philippart Post: According to the Boinc FAQ, the -185 error means: This is an error that will occur when: - if BOINC couldn't start the application - if files are missing - catch other error returns - on nonzero exit or signal - if exceeded resource limit - catch-all for resume/start errors I would detach, delete the rosetta project folder (normally located here: C:Program FilesBoincprojects) and then attach to the project again.
24) Message boards : Rosetta@home Science : Rosetta@Home Myspace Group (Message 48198) Posted 31 Oct 2007 by Tom Philippart Post: In August 2006 I created a Myspace group to promote Rosetta@Home on the social network. The idea was raised in a thread and I got the permission to use the logo from someone on this forum. Somebody asked me today wheter I'm really allowed to use the logo even though it is copyrighted. Could a moderator or a barkerlab member please confirm that I have the permission to use the logo to promote the project? Thanks PS: Here's the link: http://groups.myspace.com/rosettaathome
25) Message boards : Number crunching : 140+ times more granted then claimed (Message 47944) Posted 23 Oct 2007 by Tom Philippart Post: Now I see why my credits granted often is much less than credits claimed!! ouch! my single core can not keep up with that machine Now I see why I saw a granted credit decline.
26) Message boards : Rosetta@home Science : question du Frenchy (Message 47441) Posted 6 Oct 2007 by Tom Philippart Post: Sorry, I can't speak english. La question concerne l'approche de la recherche de l'�nergie minimale touv�e dans le repliement de la chaine d'acides amin�s. Sommes nous s�r que la situation cible est une situation d'�nergie minimale? Le repliement �tant dynamique, l'ordre de la s�quence dans le temps peut faire au mieux "au moment du repliement". La situation "interm�diaire" peut s'av�rer stable pour la suite de la proc�dure. Et le r�sultat obtenu suivant une logique dynamique du repliement pourrait ne pas �tre la situation d'�nergie minimale statiquement. Donc je repose ma question : Sommes nous s�r que la situation cible est une situation d'�nergie minimale absolue? merci Non, nous ne pouvons jamais être sûr que nous avons trouvé la situation d'énergie minimale absolue du fait que c'est un repliement aveugle. Les prédictions ne diffèrent pourtant que quelques angstroms. Pour déterminer la situation d'énergie minimale, un grand nombre de WUs pour une protéine est envoyé aux utilisateurs de rosetta@home, ces WUs rapportent les meilleures résultats comparés aus autres résultats intermédiaires de la WU. Puis les meilleurs résultats des WUs sont comparés entre eux avec les autres WUs de la protéine. Pour visaliser cela, vous pouver cliquez sur "Results" sur la page principale. Voici un exemple d'une protéine calculée l'année dernière: http://boinc.bakerlab.org/rosetta/rah_top_predictions/casp7_pictures/t330-2.jpg Cette image montre les différences minimales entre la situation minimale absolue et la situation prédite. Le plus d'ordinateurs participent à rosetta, les prédictions deviendront de plus en plus exactes.
27) Message boards : Number crunching : E2180 vs. E4500, is a 2MB L2 Cache profitable ? (Message 47375) Posted 4 Oct 2007 by Tom Philippart Post: I can't provide exact numbers, but a bigger L2 cache does have a big influence on performance in protein folding. On folding@home it nearly doubles the performance to have more than 2MB of L2 Cache.
28) Message boards : Number crunching : Every Wu ends up in an error (Message 47280) Posted 1 Oct 2007 by Tom Philippart Post: I checked a few of your results. These errors aren't due to overclocking or other tuning, these are software errors. I recommend to completely uninstall Boinc (delete the Boinc folder too, normally located here: C:Program FilesBoinc). Then reinstall the Boinc client and attach to rosetta@home. This should fix the problem. I doubt that you get new WUs today since you already reached the maximum of WUs/day on that computer. You'll have to wait till tomorrow. ;) Greetings
29) Message boards : Number crunching : x64 or x86? (Message 47174) Posted 28 Sep 2007 by Tom Philippart Post: Rosetta doesn't have a windows 64bit application (yet), the application you're talking about is in reality a true 32bit application with a wrong label in order to let people running a 64bit version of Boinc contribute to this project.
30) Message boards : Rosetta@home Science : SAMPL prediction challenge (Message 46863) Posted 23 Sep 2007 by Tom Philippart Post: Does Rosetta participate in the SAMPL challenge? SAMPL challenge URL The challenge is similar to the CASP challenge, but on a pharmaceutical level. It would be interesting to see how Rosetta performs in it.
31) Message boards : Number crunching : BOINC and Windows Vista (Message 46837) Posted 23 Sep 2007 by Tom Philippart Post: Boinc still isn't 100% Vista compatible: - If you use version 5.10.x, it won't startup automatically - If you use version 5.8.x it will start up automatically, but you have to close it manually before shutting down in order to prevent work from getting lost (only occures sometimes, it is horrible if you crunch very long WUs) The next version of Boinc: version 6 (currently in alpha testing) will fix all problems related to Vista.
32) Message boards : Rosetta@home Science : ferredoxin-like WUs (Message 46779) Posted 22 Sep 2007 by Tom Philippart Post: These new WUs sound interesting, what are they used for?.. For the moderators, this comment of course goes hand in hand with the comments about updating on a regular basis the Active Work Units Log thread. ;) Maybe we'll get some more attention this way. ;)
33) Message boards : Rosetta@home Science : ferredoxin-like WUs (Message 46762) Posted 22 Sep 2007 by Tom Philippart Post: These new WUs sound interesting, what are they used for? The WU info only says: "denovo structural design: ferrodoxin-like fold". Thanks in advance
34) Message boards : Number crunching : Why no open source/FreeBSD client version? (Message 46724) Posted 21 Sep 2007 by Tom Philippart Post: I can't find anywhere as to exactly why Rosetta at home isn't open source? I can't use it on my FreeBSD machine. I run SETI at home on it mainly because I set it so it uses my full CPU power and just let it do it's thing. I don't like windows and I don't use linux for anything other than asterisk and embedded systems. My ability to contribute to the project is limited because the client is not open source or run on FreeBSD. Jonathan You can get the rosetta source code, you have to e-mail the rosetta team to get it.
35) Message boards : Number crunching : Work unit creation (Message 46683) Posted 20 Sep 2007 by Tom Philippart Post: this idea has been discussed by a number of people in this thread: click here At that time I made a sample progress report page, much more simplified than the information Mod.Sense posted: click here I only focussed on the diseases, however any reasearch done on rosetta@home would fit in this model.
36) Message boards : Rosetta@home Science : Where we are and where we are going. (Message 44408) Posted 27 Jul 2007 by Tom Philippart Post: Sorry, I just got back from a several week family vacation a few hours ago. Will give you an update by early next week (I need to catch up on the current status of the projects as well!). Thanks in advance!!
37) Message boards : Number crunching : Error To work on no Task HELP (Message 44110) Posted 23 Jul 2007 by Tom Philippart Post: "there was work but you don't have enough disk space allocated" Dies bedeutet, dass auf der Partition, auf der sie Boinc installiert haben nicht genügend freier Speicherplatz vorhanden ist. Der windows "disc cleanup" kann ihnen behilflich sein beim befreiden des Computers von unnötigen Dateien.
38) Message boards : Number crunching : Error To work on no Task HELP (Message 44105) Posted 23 Jul 2007 by Tom Philippart Post: Good day I have the problem, which I cannot start new “WU's”, can I this manually adjust or can possibly who help itself? Excuses please for my English! And: Communication around xx: xx for a long time retarded one can set this to 0 and let several WU's run? German: Guten Tag ich habe das problem, das ich keine neuen "WU's" starten kann, kann ich dieses manuell einstellen oder kann mir irgendwer helfen? Und: Kommunikation um xx:xx lange verzögert kann man dies auf 0 stellen und mehrere WU's laufen lassen? Die Kommunikationsverzögerung kann man nicht auf 0 stellen, das wird vom Server geregelt, sie bestimmt nach welchen abständen wieder nach neuen WUs gefragt werden kann. Man kann maximal nur 1 WU pro Kern laufen lassen (Hyperthreading bei Intel ist eine Ausnahme).
39) Message boards : Number crunching : Error To work on no Task HELP (Message 44103) Posted 23 Jul 2007 by Tom Philippart Post: Good day I have the problem, which I cannot start new “WU's”, can I this manually adjust or can possibly who help itself? Excuses please for my English! German: Guten Tag ich habe das problem, das ich keine neuen "WU's" starten kann, kann ich dieses manuell einstellen oder kann mir irgendwer helfen? Ich nehme an, dass sie bereits herunter geladen wurden. Sehen sie bitte nach, ob im "Tasks" Reiter bei "Status" eine Meldung wie "Waiting to run" steht, falls nicht, kopieren sie bitte einen Ausschnitt aus dem "Messages" Reiter. Direkt kann ich ihnen leider nicht helfen, da zu wenige Informationen vorliegen.
40) Message boards : Number crunching : Benchmark error (Message 44063) Posted 22 Jul 2007 by Tom Philippart Post: don't care about it, rosetta doesn't use the benchmarks to grant credits.

Previous 20 · Next 20