Message boards : Projects : Coronavirus projects?
Message board moderation
Author | Message |
---|---|
Send message Joined: 4 Jul 17 Posts: 35 |
Hello, Are there any projects working on the coronavirus? I would like to be a part of it. Thank you |
Send message Joined: 25 Nov 05 Posts: 1654 |
|
Send message Joined: 10 May 07 Posts: 1432 |
Stanford group wants to use your computer to help researchers study the coronavirus Folding@home (A NON BOINC PROJECT) takes up the fight against COVID-19 / 2019-nCoV. |
Send message Joined: 7 Apr 13 Posts: 64 |
That is good news on Rosetta and Folding@Home, attempting to convince others to crunch that data. FYI: The world's most powerful computer just spent a few days analyzing the Covid-19 virus, obtaining results that otherwise would have taken several months.. They expect to have another go at it with a more accurate model of the SARS-CoV-2 virus. The Summit system has 220,800 CPU cores, 188,416,000 CUDA cores, 9.2PB of memory, and 250PB of mixed NVRAM/storage for the task. Summit Joins the Fight Against the Coronavirus |
Send message Joined: 8 Nov 19 Posts: 718 |
That is good news on Rosetta and Folding@Home, attempting to convince others to crunch that data. Thing is, the most powerful supercomputer isn't used for Corona, and the fah network is amongst the 10 most powerful supercomputers in the world, in part thanks to the plentyful Nvidia GPUs. The issue with boinc is that it's mostly not time based. Corona is time based, meaning they need results as fast as possible. Boinc is for those who can only crunch data a few times a week, or connect to the internet a few times a week to upload results. Fah doesn't make sense for them. Corona doesn't make sense for a boinc project, unless the deadline is like less than 24 hours; and in this case 24 hours is a long time. |
Send message Joined: 14 Aug 19 Posts: 55 |
A list of the projects we're aware of is here. Folding@Home is pushing a lot of COVID-19 work now if you can get it. Their servers are overwhelmed by all the compute power that's come online in the last few days. That's causing problems both getting and returning work, at least for GPU clients. Work assignment is also buggy, the client can get hung on a server that's not responding or issuing work and won't move on to another. I've set up exclusions in BOINC so that if my machine can't get FAH work, BOINC will resume. If you can put up with these quirks the work has run without any problems for me. Team USA forum Follow us on Twitter Help us #crunchforcures! |
Send message Joined: 27 Mar 20 Posts: 8 |
Hi I'm back to boinc/fah for corona. I was with boinc at 2008 for few years. I't pity scientific projects is so bugy and working if it like to work. And.. not open source. After few days (few servers, few gpu was working) I suspend work. I'm not 100% sure my electricity and equipment is used in correct way But is looks HUGE to see people around globe could make ew fah 767 977 TFLOPS. Compared eg Summit 148,600.0 TFLOPS and others It's pity I have some spare cpu (sometimes few gpu) to be used Best |
Send message Joined: 28 Mar 20 Posts: 1 |
"The issue with boinc is that it's mostly not time based. Corona is time based, meaning they need results as fast as possible. Boinc is for those who can only crunch data a few times a week, or connect to the internet a few times a week to upload results." That's not true. I have two machines running BOINC, and now with Rosetta, 24/7. I'm sure others do too. |
Send message Joined: 29 Aug 05 Posts: 15549 |
Corona doesn't make sense for a boinc project, unless the deadline is like less than 24 hours; and in this case 24 hours is a long time.Says the expert in everything and BOINC. I'd appreciate it if you'd start posting things like this as 'In my opinion..." If they needed the results back within 24 hours, they'd use an in-house super computer. But since the medical companies don't all have that ability, BOINC and FAH are the next best choice. Rosetta has a deadline of just 7 days on their COVID-19 tasks, which is enough. The vaccine when found won't be on the market immediately anyway, that requires a lot of testing first. |
Send message Joined: 19 Jan 18 Posts: 66 |
In addition to the previously mentioned projects (Rosetta, Folding*, https://Fold.it*,) IBM (through World Community Grid) will soon join research efforts with OpenPandemics https://www.ibm.org/OpenPandemics Are there any projects working on the coronavirus? I would like to be a part of it. Thank you * Non-BOINC |
Send message Joined: 8 Nov 19 Posts: 718 |
I just saw the notice of World community grid as well! OpenPandemics is joining IBM and WCG in an effort to find treatments against COVID 19. More info here: https://www.ibm.org/OpenPandemics Subscribe at the bottom, and you'll be informed when the new project is ready! On a side note, Folding at home only fights COVID 19 on CPU. So does Rosetta. WCG also only runs CPU-only jobs. To join World Community Grid, please visit their home page, and register! |
Send message Joined: 5 Oct 06 Posts: 5123 |
WCG also only runs CPU-only jobs.WCG runs the computing platform, but they don't provide the science. They host other people's science. If the Scripps Institute - who are the science partner for this IBM/WCG effort - code a GPU application, WCG will be ready and willing to host it. https://www.scripps.edu/news-and-events/press-room/2020/20200401-OpenPandemics-covid19.html |
Send message Joined: 8 Nov 10 Posts: 310 |
Correct, and Juan Hindo from WCG has said that they hope to get the GPU version of AutoDock (OpenCL) sometime in the future but there is no ETA. If you remember, the HCC (Help Conquer Cancer) project, a predecessor to MCM, had a GPU app. https://www.worldcommunitygrid.org/about_us/viewNewsArticle.do?articleId=210&post=-21035002_210&language=en_US |
Send message Joined: 14 Aug 19 Posts: 55 |
I...On a side note, Folding at home only fights COVID 19 on CPU.... According to the FAQ, GPU and CPU work is available, although my experience with GPU work it's almost impossible to get. I'm maintaining a current list of projects with COVID-19 work here. Less publicized projects include Ralph@Home (test project for Rosetta) and BOINC@TACC. Work for these two is spotty at best and TACC has its own problems, but they're on the list if you want to try them. Team USA forum Follow us on Twitter Help us #crunchforcures! |
Send message Joined: 5 Oct 06 Posts: 5123 |
Hopefully GPU app comes soon after launch.I'm sure apps will be developed to meet the needs of the science. Any other consideration is irrelevant. |
Send message Joined: 8 Nov 19 Posts: 718 |
I have emailed WCG, in reference to older style GPUs (mostly AMD), when they will be supporting modern Nvidia GPUs, and they replied that there is no foreseeable deadline, or if they ever will support it. From config, you can select AMD GPU, but I don't have one, so I don't know if AMD GPUs are supported. They also have some very serious tasks some running more than 1 day on a 4Ghz core. Those are the tasks you really want to assign to a GPU. |
Send message Joined: 29 Aug 05 Posts: 15549 |
They also have some very serious tasks some running more than 1 day on a 4Ghz core.Just because a task is taking so long on a CPU, doesn't mean it's capable of being run on a GPU. Not all science applications can be converted to CUDA or OpenCL (or Vulkan), due to the (proprietary) coding language used, due to memory constraints, or because internal testing has shown that it just doesn't work. The coding could probably be gotten past with a BOINC wrapper. Or by running the main part of the calculations on the CPU and only some on the GPU. Memory constraints, most GPUs still have only got 4GB of memory, some have 8GB, 11GB or 16GB. If you want anything more, you're talking several thousands or dollars/euros worth of dedicated co-processors, something projects won't ask of users. The tasks run have to be converted to kernels that can be run on the GPU. Not everything can so easily be converted this way. Or when done it takes up way more than the memory amount of the GPU. Or it can only be done in Linux, for which you then need a VM to run it on other operating systems, but the VM makes it that the GPU cannot be used. As for multi-threaded OpenCL, this can be run on the CPU as well. You can ask the project that they release an OpenCL application for CPU, which takes up all the cores of the CPU to calculate the task. Milkyway does this with their NBody application, which is multi-threaded (but not necessarily OpenCL). |
Send message Joined: 25 May 09 Posts: 1296 |
And add to Jord's list - that the project has access to developers with the appropriate skills. |
Send message Joined: 23 Feb 08 Posts: 2489 |
Some problems are linear. The result of the last step is needed for the next step. They can't be run in parallel, so they are single core tasks. Now that doesn't mean you couldn't run that task on a GPU, but only one core of the GPU would be used, the rest would sit idle. |
Send message Joined: 8 Nov 10 Posts: 310 |
BOINC should just put up a sticky (in large, black letters): Why All Projects Do Not Use GPUs It would save a lot of repetition. |
Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.