Help in the fight against COVID-19!

Message boards : News : Help in the fight against COVID-19!

To post messages, you must log in.

Previous · 1 · 2 · 3 · Next

AuthorMessage
Profile robertmiles

Send message
Joined: 16 Jun 08
Posts: 980
Credit: 10,685,613
RAC: 4,118
Message 93267 - Posted: 3 Apr 2020, 19:10:07 UTC - in response to Message 93264.  

How can I join my computer cycles to the covid19 project?

Under Rosetta@Home, you can't choose only work related to COVID-19, but this is the only BOINC project I've found where some of it is.

They give such work a higher priority than their other work, but are running short of new work often enough that all of their work eventually gets done.

Just follow the instructions for participating in Rosetta@Home to do what you can.
ID: 93267 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Applejacks

Send message
Joined: 20 May 17
Posts: 3
Credit: 10,714,773
RAC: 0
Message 93268 - Posted: 3 Apr 2020, 19:13:48 UTC

Thanks for the update !
ID: 93268 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Falconet

Send message
Joined: 9 Mar 09
Posts: 101
Credit: 223,483
RAC: 1,404
Message 93278 - Posted: 3 Apr 2020, 19:50:32 UTC
Last modified: 3 Apr 2020, 19:52:03 UTC

The Rosetta/L3 cache issue has been discussed quite a lot over at the Microbiome Immunity Project (MIP) forums on World Community Grid, a project which also uses Rosetta.
Running more than 1 WU per every 5 MB of L3 cache resulted in increase runtime for all tasks running (non-Rosetta too), which results in loss of efficiency considering the increase in runtime and constant electricity usage. One of the MIP scientists said they were aware of the issue and were looking into it, but this was years ago by now.
ID: 93278 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile pembo
Avatar

Send message
Joined: 11 Oct 16
Posts: 2
Credit: 516,933
RAC: 0
Message 93291 - Posted: 3 Apr 2020, 21:51:33 UTC

Good stuff, many thanks!
ID: 93291 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Hepworks

Send message
Joined: 22 Feb 09
Posts: 1
Credit: 95,805
RAC: 0
Message 93292 - Posted: 3 Apr 2020, 21:52:24 UTC - in response to Message 93153.  

Hi! Thank you so much for this project. Where is the download link to update? I'm having trouble finding it.

John Hepworth
ID: 93292 · Rating: 0 · rate: Rate + / Rate - Report as offensive
a0062995

Send message
Joined: 10 Mar 20
Posts: 3
Credit: 675,004
RAC: 4
Message 93293 - Posted: 3 Apr 2020, 21:53:27 UTC - in response to Message 93278.  
Last modified: 3 Apr 2020, 22:01:58 UTC

The Rosetta/L3 cache issue has been discussed quite a lot over at the Microbiome Immunity Project (MIP) forums on World Community Grid, a project which also uses Rosetta.
Running more than 1 WU per every 5 MB of L3 cache resulted in increase runtime for all tasks running (non-Rosetta too), which results in loss of efficiency considering the increase in runtime and constant electricity usage. One of the MIP scientists said they were aware of the issue and were looking into it, but this was years ago by now.


I don't think so.
I have one E5-2690 (8 Cores @ 3.3 GHz with 20 MB L3 Cache), one E5-2670 v3 (12 Cores @ 2.6 GHz with 30 MB L3 Cache), one i5-4570S (4 Cores @ 3.2 GHz with 6 MB L3 Cache) and all are crunching WUs with the same amount of time (~29000 seconds).
And even if i have for a time only a fraction of cores running WUs, it would not get faster (if correction for turbo is factored in).
The last 6 WUs on my E5-2670 v3 are run simultanously, they are not faster than the other ones that ran 12-parallel.
So could you give some examples?
ID: 93293 · Rating: 0 · rate: Rate + / Rate - Report as offensive
JoshuaScholar

Send message
Joined: 26 Mar 20
Posts: 18
Credit: 232,183
RAC: 0
Message 93295 - Posted: 3 Apr 2020, 22:07:22 UTC - in response to Message 93265.  

You can program non-temporal accesses to memory (bypassing the cache) in assembly language but that never seems to improve anything.
ID: 93295 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Falconet

Send message
Joined: 9 Mar 09
Posts: 101
Credit: 223,483
RAC: 1,404
Message 93302 - Posted: 3 Apr 2020, 22:35:03 UTC - in response to Message 93293.  
Last modified: 3 Apr 2020, 22:38:37 UTC

The Rosetta/L3 cache issue has been discussed quite a lot over at the Microbiome Immunity Project (MIP) forums on World Community Grid, a project which also uses Rosetta.
Running more than 1 WU per every 5 MB of L3 cache resulted in increase runtime for all tasks running (non-Rosetta too), which results in loss of efficiency considering the increase in runtime and constant electricity usage. One of the MIP scientists said they were aware of the issue and were looking into it, but this was years ago by now.


I don't think so.
I have one E5-2690 (8 Cores @ 3.3 GHz with 20 MB L3 Cache), one E5-2670 v3 (12 Cores @ 2.6 GHz with 30 MB L3 Cache), one i5-4570S (4 Cores @ 3.2 GHz with 6 MB L3 Cache) and all are crunching WUs with the same amount of time (~29000 seconds).
And even if i have for a time only a fraction of cores running WUs, it would not get faster (if correction for turbo is factored in).
The last 6 WUs on my E5-2670 v3 are run simultanously, they are not faster than the other ones that ran 12-parallel.
So could you give some examples?



Runtime at Rosetta@home is fixed, so tasks always run per the runtime preferences you set.
Faster or slower machines at Rosetta@home only do more or less work per tasks but the runtime is the same per prefs. If you set it at 8 hours, then your tasks run at 8 hours on your Xeons. If I were to set 8 hours for the tasks on an old Pentium 4, Rosetta tasks would still run for the same 8 hours but do much less work than your Xeons.

Doneske posted a better post than mine, including the MIP Scientist response on WCG - https://boinc.bakerlab.org/rosetta/forum_thread.php?id=12554&postid=93283#93283
ID: 93302 · Rating: 0 · rate: Rate + / Rate - Report as offensive
a0062995

Send message
Joined: 10 Mar 20
Posts: 3
Credit: 675,004
RAC: 4
Message 93305 - Posted: 3 Apr 2020, 22:58:29 UTC - in response to Message 93302.  
Last modified: 3 Apr 2020, 23:18:12 UTC

So how do you find out it is running less efficent, if the WU is always running 8 hours?
How do you know how much "work" you have done for your WU?

I have made these 15 second intervall measurements (Idle Desktop, 1 COVID-19 rosetta, 8 PPSE LLR primegrid WUs, one of those finding a prime i take for comparison measurements of energy consumption, speed and achievable turbo, got quite a lot data for several CPUs by now)





Can't see a lot more misses, on either situation, but am starved for now regarding rosetta WUs.
ID: 93305 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Grant (SSSF)

Send message
Joined: 28 Mar 20
Posts: 685
Credit: 3,941,033
RAC: 22,922
Message 93310 - Posted: 3 Apr 2020, 23:19:24 UTC - in response to Message 93305.  
Last modified: 3 Apr 2020, 23:20:26 UTC

So how do you find out it is running less efficent, if the WU is always running 8 hours?
In your Account page, Computers, Details, Application details, Average Processing Rate (APR).
The number of Floating Operations performed each second. The higher, the better (ie, the more numbers crunched per hour, day, week etc).
So far it's looking less efficient, but only by a small amount- about 0.6% at this stage (at least on my system- those with more cores & threads the impacts will probably be greater). However, we would need to run the same Tasks that were processed using the other Applications on this Application for a true comparison- the present Tasks may be "chewier" than the previous ones, and may have run even slower on the previous Application than they do on this new one.
Grant
Darwin NT
ID: 93310 · Rating: 0 · rate: Rate + / Rate - Report as offensive
simonc

Send message
Joined: 3 Apr 20
Posts: 1
Credit: 0
RAC: 0
Message 93311 - Posted: 3 Apr 2020, 23:24:44 UTC
Last modified: 3 Apr 2020, 23:25:04 UTC

I switched my pi farm across to rosetta@home after using it for seti@home for the past year, as I was hoping to contribute to the project, have got a “no work available to process “ message for the last 10 hours or so....normal for there to be fewer WU?
ID: 93311 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Grant (SSSF)

Send message
Joined: 28 Mar 20
Posts: 685
Credit: 3,941,033
RAC: 22,922
Message 93313 - Posted: 3 Apr 2020, 23:34:31 UTC - in response to Message 93311.  

I switched my pi farm across to rosetta@home after using it for seti@home for the past year, as I was hoping to contribute to the project, have got a “no work available to process “ message for the last 10 hours or so....normal for there to be fewer WU?
For now, yes. Demand for work has increased by over 500% over the last few weeks.
Grant
Darwin NT
ID: 93313 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Falconet

Send message
Joined: 9 Mar 09
Posts: 101
Credit: 223,483
RAC: 1,404
Message 93315 - Posted: 3 Apr 2020, 23:48:27 UTC - in response to Message 93305.  

In my initial post, I referred to WCG. WCG is another project which runs 1 Rosetta based project, the MIP project.
Running more than 1 task per 5 MB of L3 cache of the MIP project in WCG causes other MIP tasks and non-MIP tasks to take significantly longer to finish. This I have seen as have many other users in the WCG and was acknowledged by that projects Scientist team, as per the post I linked previously.


Regarding Rosetta@home, considering runtime is fixed, I suppose one way to check the work done would to check each task log and see how many decoys it computed, for example, running all cores. And then run 1 task per 5 MB of L3 cache and attempt a comparison. But different tasks might not give accurate results.

I have only 1 MIP task running at any time on any of systems and I will probably do the same with Rosetta, which is already limited to 4 on the Ryzen due to memory constraints.


User Aurum posted this above - 2. I'm laboring under the assumption that Rosetta code incorrectly programmed the use of the L3 cache and will bottleneck if torsioo many Rosetta WUs are running simultaneously. I limited my use to one WU per 5 MB of L3 cache. Exceeding that limitation slows the entire CPU over 60%. This means that my fleet of Xeon E5s are running Rosetta at less than 20% capacity. Has this been fixed and is it safe to run full force now???

Who knows, maybe this new version is better coded.


I actually went to the WCG thread and re/read the scientist reply on the issue and he actually mentions 4 MB of L3 cache as being ideal, otherwise - A single instance of the program fits well in to a small cache. However, when you begin to run multiple instances there is more contention for that cache. This results in L3 cache misses and the CPU sits idle while we have to make a long trip to main memory to get the data we need.
ID: 93315 · Rating: 0 · rate: Rate + / Rate - Report as offensive
a0062995

Send message
Joined: 10 Mar 20
Posts: 3
Credit: 675,004
RAC: 4
Message 93317 - Posted: 4 Apr 2020, 0:03:59 UTC - in response to Message 93310.  
Last modified: 4 Apr 2020, 0:18:58 UTC

So how do you find out it is running less efficent, if the WU is always running 8 hours?
In your Account page, Computers, Details, Application details, Average Processing Rate (APR).
The number of Floating Operations performed each second. The higher, the better (ie, the more numbers crunched per hour, day, week etc).
So far it's looking less efficient, but only by a small amount- about 0.6% at this stage (at least on my system- those with more cores & threads the impacts will probably be greater). However, we would need to run the same Tasks that were processed using the other Applications on this Application for a true comparison- the present Tasks may be "chewier" than the previous ones, and may have run even slower on the previous Application than they do on this new one.


Those values do not quite make much sense.
Is the value stated per Core?
Ryzen 7 3700X 2.87 GFLOPs (8x @3.6 GHz; 30 MB L3; 51.2 GB/s RAM) (okay, old program with 32-bit programm in parallel)
E5-2690 2.74 GFLOPs (8x @3.3 GHz; 20 MB L3; 51.2 GB/s RAM)
i5-4570S 2.75 GFLOPs (4x @3.3 GHz; 6 MB L3; 25.6 GB/s RAM) (only 3 values so far)
E5-2670 v3 2.79 GFlops (12x @2.6 GHz; 30 MB L3; 59.7 GB/s RAM)

If the L3-cache is a bottleneck especially the i5-4570S should be a lot slower it the value is per core.
And the Ryzen should be a lot faster regarding the fact it is newer and should be way more efficient.

I've sent an email to that Doug Refren, maybe he could elaborate a bit more.

I am also open for testing, if you could tell me which project in WCG uses fixed work, not fixed runtime I am happy to run tests for comparison.
ID: 93317 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Bernardo Garcia Loaiza

Send message
Joined: 25 Mar 20
Posts: 2
Credit: 3,691
RAC: 0
Message 93819 - Posted: 8 Apr 2020, 5:58:25 UTC - in response to Message 93264.  

I suppose by default I am doing it, just install Rosseta and that is (?)
ID: 93819 · Rating: 0 · rate: Rate + / Rate - Report as offensive
nnneal

Send message
Joined: 11 May 19
Posts: 35
Credit: 5,547,148
RAC: 11,468
Message 93864 - Posted: 8 Apr 2020, 13:19:11 UTC - in response to Message 93819.  

I suppose by default I am doing it, just install Rosseta and that is (?)


Most if not all of the Rosetta tasks I get lately are covid19-related. So if you are receiving credit then you are contributing to covid19.

Neal
ID: 93864 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile bitsonic

Send message
Joined: 21 Mar 20
Posts: 9
Credit: 1,530,709
RAC: 6,172
Message 93870 - Posted: 8 Apr 2020, 15:06:10 UTC

Greetings from China and big thank you to the team!I have been with SETI@home since 1999. Now I take my computers doing another meaningful thing here. My country has been suffering with this virus. I donate all my computer power here. I hope we will help scientists finding vaccine eventually.
ID: 93870 · Rating: 0 · rate: Rate + / Rate - Report as offensive
chrislaf

Send message
Joined: 28 Mar 20
Posts: 3
Credit: 1,487,482
RAC: 0
Message 93880 - Posted: 8 Apr 2020, 16:35:01 UTC

So, am I supposed to update my client to participate in this new capability? Kind of confused as to whether this is a call to action or just a status update.
ID: 93880 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Gh0sth4nd
Avatar

Send message
Joined: 10 Mar 20
Posts: 1
Credit: 109,220
RAC: 0
Message 93883 - Posted: 8 Apr 2020, 16:49:49 UTC

This is good news
COVID-19 we comin!

I am a little bit proud to be able to do my part to help finding a solution, even if its only a very small part.
ID: 93883 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Cliff

Send message
Joined: 29 Mar 09
Posts: 1
Credit: 23,427
RAC: 0
Message 93884 - Posted: 8 Apr 2020, 17:12:45 UTC

are tasks called Rosetta 4.12 related to Covid-19?
ID: 93884 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Previous · 1 · 2 · 3 · Next

Message boards : News : Help in the fight against COVID-19!



©2020 University of Washington
https://www.bakerlab.org