Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
![]() |
World Community Grid Forums
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 71
|
![]() |
Author |
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7615 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Tech query: What is the best CPU for CEP2? Is there a breakdown of hosts running this project? What is the optimal hardware combo? The inverse of that query: if I have a given CPU, for what projects is it best suited? Is there a chart which offers such guidance? Thanks I can not directly answer your question, but I can tell you the best OS for running any of the projects using VINA(currently only a portion of FAAH) is any Linux variant. Generally speaking, the higher the floating point capability of your cpu, the faster it will crunch. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
It is probably a good time for extensive maintenance activities ! ... Yves And maybe building a new rig. ![]() ![]() |
||
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
KerSamson I fully approve
----------------------------------------![]() I agree with many comments here. Since I have business experience with some large organisations/companies, I feel constantly the typical communication dilemma of big US companies. In order to ensure that nothing wrong could be communicated the communication becomes rare and useless. It is clear that it is necessary to clearly manage (and sometimes to restrict) the communication. However, since we are contributing to public projects, with public domain results, since the members finally spend, over the years, a lot of money, I think that it is not too demanding to ask for a more transparent and reliable communication. I can remember that, many years ago, some people mentioned the necessity for WCG to involve some contributor/member representatives in order to avoid a kind of autism from "WCG management and staff" to the members. If WCG should have a living and rich future, I think that the partnership between members and WCG management should be strongly improved. That is called "Governance"! I would like to come back to my previous post. Graphical communication is a wonderful and efficient way to share simply information, progress, challenge, or result without spending too much time, without having to say too much. The success of SekeRob's graphical updates showed how important and useful such communication could be for the members. We can think about some dashboards for informing about the progress of the project pre-release activities, about the operational troubles and progress, and finally about the assessment phase of the results by the scientists. I am still believing that we do a useful job with our contribution and participation to WCG. Nevertheless, I can easily imagine that less convinced people do not interpret in a positive way the lack of feedback after project completion. I can imagine that some people have the feeling that, finally, our contributions are not really useful and since scientists are not able to provide accurate feedback about the use of our computational contributions, those were useless. Since I worked for companies performing research and development for life science, I know very well how long the way is from the basic research to a final product over every late research and development steps. It takes usually years, sometime decades, between the beginning of a project and the final successful result. However, like in the industry, there are some milestones to achieve for deciding if an initial research project remains interesting and valuable or if it is necessary to stop spending time and money on it. Not being successful on a scientific level is not dramatic, but the contributors need to know about it. Being able to recognise that some research / investigations do not conduct to a successful product is important and valuable because it will avoid that other researchers will waste their time and resources with this approach. I believe that the support WCG (members) provides to the scientist is crucial for the future. But it is important too to maintain a living and meaningful communication. Good luck, Yves ![]() |
||
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1671 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
What is the best CPU for CEP2? Is there a breakdown of hosts running this project? What is the optimal hardware combo? Over the year crunching for WCG, I dreamed of a tool providing a dashboard regarding computation efficiency, project affinity, and failure monitoring. I started a first proof of concept last year with two students (Master project). I am just initiating a consolidated implementation with two new students. Hopefully, they will provide to me a great Christmas gift at the year end ! ... As soon as I will have a reasonable result, I will share it with you. Evaluating accurately project affinity is not so easy because several factors could have a significant impact: CPU, Operating System, concurrent projects as well. In the past, we did experience that two particular projects together on the same CPU could drop down the CPU performance. Usually, we can say that Linux x64 bring significantly better performances than Windows systems. Likewise, Windows does manage memory really inefficiently compared to Linux. In the past, I observed that Xeon (5345) and AMD (Phenom II x6) were particularly good with HCC and HCMD2. CEP2 could be a good candidate too for such CPUs. You can perform your own analysis using LibreOffice Calc or Excel. However, an "on-line" (at least daily) reporting would be really useful. Yves |
||
|
jhindo
Former World Community Grid Admin Joined: Aug 25, 2009 Post Count: 250 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
There's been a lot of discussion here about the new cancer project and why beta hasn't started. I'd like to confirm that we are in (hopefully) in the final stages of alpha testing. We had encountered a non-trivial issue earlier during alpha testing, which has now been resolved. With that issue fixed, we can focus on the remaining handful of issues. So while we're not there yet, we're getting much closer.
I also want to take this opportunity to explain a little about this process and why it's hard to estimate accurately (thus the lack of specificity about a launch date). Each research team specifies what software application is best suited to their scientific calculations. Sometimes, the researchers will pick a software that we have already integrated into the Grid (e.g. the Vina docking software is used by several of our projects), which makes the process on onboarding that project significantly easier. For this new cancer project however, the research team has developed their own custom-built application. This does mean incorporating brand a new software tool into World Community Grid. This involves an extensive process of conducting a security review of the code, ensuring compliance with any relevant open source constraints and licenses, adding checkpointing capabilities, developing the graphics for the screensaver, testing out the mechanism for work unit generation and packaging, ensuring the validity of the results generated, etc. We often come up against all kinds of surprises and unexpected issues during that process which we have to resolve before releasing the app into beta testing. I hope this helps explain why with so many variables, it becomes hard to communicate accurate timescales. Glad to say the bulk of that work for the new cancer project is behind us as we wrap up what are hopefully the last few issues. So hang in there, not long to go now before we're ready for beta! |
||
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2979 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Excellent news jhindo - and thanks for expanding on what goes on behind the scenes, as I'm sure there are many who don't actually realise what it all entails.
----------------------------------------Yes, at times we're all a little impatient, although it's far better to ensure that a new project gets off onto the right foot, than has to be taken offline once it's launched so as to fix something that was skipped in order to get it 'out there' quicker. ![]() |
||
|
cjslman
Master Cruncher Mexico Joined: Nov 23, 2004 Post Count: 2082 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Thanks for the info jhindo. This is exactly the kind of information that the crunchers need to know to keep the impatience at bay
----------------------------------------![]() CJSL Gotta keep crunching... there's a world to save !!! |
||
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Tech query: What is the best CPU for CEP2? Is there a breakdown of hosts running this project? What is the optimal hardware combo? The inverse of that query: if I have a given CPU, for what projects is it best suited? Is there a chart which offers such guidance? Thanks The following tables should contain the data to answer your question. Only combinations of os_name/p_model are included where there were at least 3 devices returning results for each of the projects included: credit_per_hour_faah credit_per_hour_fahv credit_per_hour_cep2 os_name p_model |
||
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
credit_per_hour_faah credit_per_hour_fahv credit_per_hour_cep2 os_name p_model |
||
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
credit_per_hour_faah credit_per_hour_fahv credit_per_hour_cep2 os_name p_model |
||
|
|
![]() |