Aller au contenu
PcPerf.fr

PcPerf bot

PcPerfonaute
  • Compteur de contenus

    399
  • Inscription

  • Dernière visite

Tout ce qui a été posté par PcPerf bot

  1. We've been experiencing a great deal of stats instability, due to the new advmethods projects announced recently. The problem is that these WU's are just too short for the stats system to handle, flooding the system with 10x more WU's than it's used to handling. That means that instead of completing an update in 1 hour, it now takes 10hours, which is a disaster. We're working to correct this, but it will likely take a little time to get through the backlog that's been generated. Voir l'article complet
  2. We've been experiencing a great deal of stats instability, due to the new advmethods projects announced recently. The problem is that these WU's are just too short for the stats system to handle, flooding the system with 10x more WU's than it's used to handling. That means that instead of completing an update in 1 hour, it now takes 10hours, which is a disaster. We're working to correct this, but it will likely take a little time to get through the backlog that's been generated. Voir l'article complet
  3. We've teamed up with Simbios (an NIH Roadmap center at Stanford) to start making key technologies in Folding@home available for others to use. Our protein folding work has been named one of Simbios' key Driving Biology Problems (DBPs) and there's now a protein folding DBP page. Our hope is that our developments which have been so useful in Folding@home can help accelerate the work of others as well. Right now, there are two key areas we've been working to distribute, with more coming along the way. First, we're working to make our GPU code available to others. This code will be distributed in a couple of forms. First, we'll give out a GPU-enabled version of Gromacs (basically, a standalone version of the GPU2 core), which will enable others to get major speed increases from GPUs. Next, we are working to release a GPU-enabled library (OpenMM), which will allow others to integrate GPU code into their programs. OpenMM is special in that it is a place for integrating both application developers as well as GPU vendors; much like OpenGL, our hope is that hardware acceleration vendors will now have a single API to accelerate and people who want to write applications will have a single, hardware accelerated API to use that would work on a variety of platforms. Second, we're also starting to make large data sets from Folding@home available to others. You can see some of the first data sets on this project page, and we expect to put more data up as time goes on. Folding@home donors have generated wonderful data sets that aren't possible to generate by other means, and our hope is to publish them so that other scientists can data mine them for other purposes as well. Finally, we expect to release more technologies from FAH. In particular, we will be releasing some of the key server-side algorithms which allow FAH to use lots of processors to tackle single complex problems. This will allow people to run massively parallel code on large clusters to do tasks similar to FAH (albeit on a smaller scale). Voir l'article complet
  4. In the near future, we will be releasing some new projects which require a very rapid turn-around time. These are peptide fragment simulations which we are interested in simulating for a time-sensitive collaborative project involving protein structure prediction. These WUs will go directly to the classic clients running with -advmethods. Non-classic clients (eg SMP, GPU, PS3) will not be affected, as all of these calculations will be run via the AMBER core and only the classic client supports the AMBER core. To reward users for participating in this exciting project, we will be giving a x1.5 bonus in the points awarded. What's the catch? These projects will be less rigorously beta tested, so there will be an increased risk of Early_Unit_End errors. We believe the risk of this is minor (there will likely be a higher rate of early unit ends, especially very early in the WU, but we do not expect client machines to become significantly less stable). However, if you do not wish to participate in this project, just remove the -advmethods setting from your client. This will not last forever and will likely go back to normal -advmethods usage in a few months. With that said, we do plan other uses for -adv in the future. Note: the text was updated to clarify that this is classic clients only. Voir l'article complet
  5. Several people have asked about visualization with GPU2. That is in the works and should be pretty close (not months, but hopefully weeks, maybe faster). The visualization will be real time like the PS3 (although perhaps not quite as slick, as the Sony team put a lot of effort into lots of nice small touches in the PS3 client visualization). Voir l'article complet
  6. There have been some misunderstandings on how the GPU2 core works. In particular, for small proteins like villin on GPU's with large number of stream processors (SP's) like the 3850 or 3870, the protein is too small to use a larger number of SP's unless the CPU is very fast. Some people have guessed that there is some internal SP limit. This is incorrect; the problem is that small proteins can't be parallelized amongst a large number of SP's. We are working to release larger proteins (about 2x the number of atoms) as they are more interesting scientifically and use the GPU's (even the high end ones) much closer to 100%. The exciting part for us is that the larger proteins run at almost the same speed as the slower ones on GPU's (whereas on CPU's, they're 4x slower); this is where the GPU2 code should shine. In parallel, Mike Houston at AMD is working to optimize CAL such that it has lower CPU overhead. For now, we're pushing out villin WU's as a test (good to know that the code is working well), but we expect the larger WU's to be going out soon (say a week or two, pending internal testing). Voir l'article complet
  7. The GPU2 code beta test is going well so far. While there are some issues to deal with, it looks like this client is significantly easier to use than GPU1. Also, we're excited since the science in GPU2 makes a big step forward. There's been lots of suggestions (tweaks to the client, a visualizer, PPD change), and we've been compiling lists of requests and going over what makes sense. Right now, I'd say that everything is on the table, but some suggestions are easier than others. We'll prioritize on the easy & important additions first of course (and the hard & less important changes last). Also, we're working on a visualizer that will give real time updates like the PS3, as well as a console version of the GPU2 client. We may hit a show stopper, but that's on our roadmap. We're excited about the turnout so far and are looking forward to doing exciting work together with this new technology. Voir l'article complet
  8. We're (finally) ready to start the QA process outside Stanford & ATI for the next generation GPU client (nicknamed FAH GPU2). We've gone through a limited alpha and a larger closed beta test, and we think the client is now ready for an open beta, i.e. a public release of the software that's still in an early stage. That's not to say that it's flawless, but there don't appear to be any "show stoppers" so far. If you have a 2xxx or 3xxx card, please feel free to check out the new GPU client? You can download it from here http://www.stanford.edu/~beberg/Folding@ho...00-systray.zip and a FAQ is here "http://folding.stanford.edu/English/FAQ-ATI2" We will next put this on our main download page, assuming that this stage of beta testing looks good. Please keep in mind that this is beta software. The client/core may be unstable (although it looks ok so far), the points benchmarking may change due to software or policy changes, etc. With that in mind, we'd appreciate any comments you may have -- please post them in this forum section. Thanks! Voir l'article complet
  9. We've internally dubbed the new GPU client/core "GPU2". Here's an update. The GPU2 closed beta testing has been moving along well. There are a few bugs which have shown up, but nothing we'd call "showstoppers" just yet. The beta test is outside of Stanford and ATI now. If all looks good, we plan to next have a completely open beta. Voir l'article complet
  10. We've found a bug in our server code related to SMP (and GPU2 clients, but this is just a small alpha test right now) and are working to fix it. We think we have found the problem and so we've been working to do a series of manual point recredits before putting in the new working code. If you're running an SMP client, you may be seeing some new recredit points today and maybe later this week as well. Voir l'article complet
  11. It looks like our hosting company is having problems with the network connectivity with our server. They tell us that After detecting that there is a problem with the network connectivity to some of our servers we have run extensive verifications and checks to establish the nature and the severity of the issue. It turned out that there is an unexpected network failure in our data center that affected your host server. Network failures like this are rare but unfortunately totally unpredictable. However, we would like to assure you that our server administrators have all the needed expertise and follow strict procedures to minimize the negative effect of such an event. We are now working together with the technicians from our data center on fixing the network problem as quickly as possible. For now, it seems there is a major problem somewhere along the route to some providers around the world, however we will keep you updated on the progress of our work. We'll keep donors up to date on this blog with info as we know it. Voir l'article complet
  12. We've moved the second generation GPU client to a closed beta test, outside of Stanford and ATI. We're doing just the first few stages of this, but at least everything looks ok so far. We expect to find several rough edges to fix over the next few days/weeks, but I think we're getting close (assuming no show stoppers appear). Traduction: Nous avons passé la seconde génération du client GPU en beta test fermé, en dehors de Stanford et ATI. Nous ne sommes qu'aux premières étapes de celui ci, mais pour le moment, tous semble bon. Nous nous attendons à trouver plusieurs problèmes ces prochains jours/semaine, mais je pense que nous arrivons au bout de la beta fermé (en espérant qu'il n'y ais pas de nouveaux obstacles). Voir l'article complet
  13. The Windows SMP client is a pretty important new client for us for two reasons. First, multi-core CPUs are becoming more and more common, and second, Windows clients are the most common in FAH by far. The SMP client appears to be working quite well on linux and OSX, and this is due to the fact that these machines are unix-based, and so multicore libraries (eg MPI) are much more robust on these platforms. The issues we've been seeing on Windows can be directly attributed to MPI issues. We're the first distributed computing platform to roll out MPI calculations (typically reserved for supercomputers), so we are dealing with some major growing pains issues there. To help fix this, Peter Kasson has been working hard on a new approach to improve the Windows SMP core stability. The trick was to use an improved MPI library. So far, it looks like it is indeed a significant improvement, but we're still testing it to be sure. If the Windows SMP core is not working well for you, we hope that the new fixes will help. For now, if the SMP client isn't working well for you, one can always run multiple regular Windows clients and still make major contributions to FAH (and lots of extra points than a single client!). You can see more info on our Windows console FAQ. Finally, we're working to reintegrate and revamp our Windows clients in general. We've revamped the graphics, made it much more stable (and better looking), and we've set up a single client to serve as both regular and SMP clients (with GPU functionality on the way). We're excited about these new developments and donors will be seeing these advances in the next few weeks, if all goes well in our beta tests. Voir l'article complet
  14. Several people have asked about FAH's stance to open source and whether we'd open up source for parts of FAH. We have a FAQ discussing this at http://folding.stanford.edu/English/FAQ-OpenSource Of note is our opening up of FAH GUI client display. Our client OpenGL kit is now available. The purpose of this is for OpenGL programmers to be able to us make graphic modes for the Folding@home client, and viewers. If you are an OpenGL programmer, download the .zip file, and read the README.TXT for instructions. Visit the forum to discuss, ask questions, and show off your work. http://www.stanford.edu/~beberg/FAH-OpenGL-Kit-1.zip Voir l'article complet
  15. We're starting to broaden our beta test for the GPU client. You may from time to time see 27xx class project numbers appear on the Project Summary page -- these are the new GPU test projects we're running. So far, so good. We're working with ATI to work out a clean way to bundle what we need for the new client, finishing up some tests at ATI, and then we should be in good shape for a more broad (but still closed) GPU beta test. Voir l'article complet
  16. We're pretty happy with the results of the QA from our new GPU core and we are working with ATI to do some more extensive QA. If that goes well, the next step is a closed beta, and then an open beta (i.e. downloads on our web site). It's taken a while to get to this point due to many fixes and changes under the hood. Some of these changes have to do with new ATI development tools for using GPUs (these changes will help the reliability of the code as well as solve a lot of issues donor side, such as problems with drivers) as well as new science we've put into the GPU code, to bring it up to date with what we now do on the PS3 (in fact, some features of the GPU code are not present in the latest 1.3.1 PS3 version, but we hope they will be included in a future version). I'll keep everyone posted here as we go along. It's hard to tell how long this current QA stage will take (and a lot of it depends on whether any major bugs surface), but we're happy to see this baby go out the door, at least with its first small steps. Voir l'article complet
  17. Sony has released an upgraded client for the PS3. Check out Noam Rimon's (he's the lead developer at Sony) post: Hello, Starting today a new version of Folding is available. This version is a minor fix to the previous 1.3 version and adds a better tuned algorithm that handles peak performance hours of Folding@home network, by acting differently to certain network errors if those occur. It is recommended that you update to the new version by quitting the application and restarting it. Your current WU will not be damaged in any way, in-fact Folding will continue from the exact point you have stopped it. Please also check the official Sony Folding site here: http://www.scei.co.jp/folding/en/update.html We can already see a stabilized network activity for the last 5 days. With the new update we are certain that the network will continue to be stable for good. At this point I would like to thank people again for being patient and continue to donate to our project while Stanford & Sony was working on the fix. Thanks, Noam Sony Development Team The upgrade should significantly improve the networking performance. Voir l'article complet
  18. It looks like the system has stabilized a bit, although there is still lots of room for improvement (that will come especially with a new client). For now, here 's an update on our performance: it looks like the PPD on the PS3 donor name has settled down back to the levels we saw before this mess. We will continue monitoring and trying to see what we can do to improve wait times (wait times for WU's can still get long), but so far so good! Voir l'article complet
  19. The School of Medicine Networking team has scheduled a brief network outage for one of our main server rooms, projected to be 2-15 minutes in length, for 6 pm on Thursday, February 14, 2008. With backup AS's and work servers, we should be fine (especially since this will be so brief). Voir l'article complet
  20. It looks like the changes we made have been helping. The production is now getting back to the pace we had before all of this mess started. I don't want to get too excited about this just yet, but I think we may be in ok shape until the Sony client comes out. Here's a summary of production for the PS3 (default) donor. Voir l'article complet
  21. We're seeing signs that are most recent modifications are helping and we may have the PS3 network of machines back on track. Nevertheless, we're still working to see how we can improve it from here (there's lots of aspects that are non-ideal) before the client revision from Sony. Voir l'article complet
  22. Sony has been working on a new client to fix several client-side bugs which have come to light during the last few days. Once that new client is in, we can set the servers back to normal and all should be well, assuming the client addresses the issues we're facing. For now, we (Stanford) have been trying to do what we can server-side to work around some critical bugs in the 1.3 client to allow for both client uploads and downloads. It's easy to get uploads or downloads working, but with the current state of 1.3, it's hard to get both. We have had to rewrite server code to work around these 1.3 bugs and I think we've made some progress. The bugs in the 1.3 code are very subtle and the sort that only comes up once a critical mass of 1.3 clients exist and start to hit the server collectively in incorrect ways. Sony QA didn't show this up, as one can't QA 30,000 clients. However, we have discussed in detail with Sony some ideas for helping to prevent this in the client in the future, and this behavior is being removed for the upcoming patch. I want to address some misconceptions in the comments posted here. There is a problem in the client, and the server changes were made to address this and work around the bug (the server changes did not create the problem, but were done in response to the problem). Also, Sony QA is extremely rigorous, but these sorts of things can't be found until clients hit critical mass. Finally, we have tried several server code changes and the reasons they didn't help appear to be that the client had multiple aspects which didn't behave as expected (eg client bugs) which rendered our changes not as useful as one would have expected at first (forcing us to reconsider what one can do server-side). Last night, we tried a new strategy where we use the AS to help control the weight of clients going to the WS. This seems to be working as we're getting lots of uploads right now as well as a steady stream of downloads. We hope that this is a reasonable balance and if not, we will continue to see what we can do from here until the new client comes out. Voir l'article complet
  23. We did the test I discussed in update #11 at 6AM-7:30AM PST this morning. It was useful in debugging, but we've decided to hold off with that plan for now. Voir l'article complet
  24. We had another meeting with Sony to discuss the results of our recent server code change to improve the situation. We have a working model of the situation and we will continue to see what we can do server side before the client patch comes out. We will be testing an approach which means that many clients (99%) won't get assigns for a while, allowing the remaining 1% to get work. Once they have work, they'll go away crunching, allowing a new 1% to get work. If all goes well, all should get work without all pounding the server simultaneously. The new client will take care of this automatically (as this is already done in the non-PS3 clients), but we will handle this from the Assignment Server "manually" to take care of this until the new client gets out. Once we get over the hump, we should be ok. Voir l'article complet
  25. PS It looks like with the recent credit update, the points have done a lot to get back on track. Check out some 3rd party stats (eg http://folding.extremeoverclocking.com/use...s=&u=207511 for the latest). Voir l'article complet
×
×
  • Créer...