1000+ new hosts in past two days & question

Message boards : Number crunching : 1000+ new hosts in past two days & question

To post messages, you must log in.

AuthorMessage
Michael G.R.

Send message
Joined: 11 Nov 05
Posts: 264
Credit: 11,247,510
RAC: 0
Message 29519 - Posted: 17 Oct 2006, 16:47:37 UTC

1092 new hosts in the past 24 hours, a bit more than that the day before. Nice little spike.

Maybe this has been discussed before, and if so, please just point me in the right direction, but I was wondering if the Rosetta team had some kind of threshold that they are aiming for before rolling out some new science.

f.ex., Do we need more than 40 teraflops of computational power before it makes sense to send out some more complex proteins? Do we need X number of teraflops before a new algorhythm is used or whatever?

In short: Are we waiting for a certain number of crunching power before implementing some things that are already ready to go?

Thanks,
ID: 29519 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1830
Credit: 119,208,549
RAC: 2,278
Message 29523 - Posted: 17 Oct 2006, 17:22:10 UTC

To expand on that, I would think that it is just as important to have a certain computational capacity with enough available RAM to process some of the larger jobs that have been coming through.

I guess when dealing with just the computational power, if there isn't 'enough' then it will just take longer to reach the target.
ID: 29523 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Michael G.R.

Send message
Joined: 11 Nov 05
Posts: 264
Credit: 11,247,510
RAC: 0
Message 29526 - Posted: 17 Oct 2006, 17:43:13 UTC - in response to Message 29523.  

Good point.
ID: 29526 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Astro
Avatar

Send message
Joined: 2 Oct 05
Posts: 987
Credit: 500,253
RAC: 0
Message 29527 - Posted: 17 Oct 2006, 17:46:47 UTC

There have been decent increases in user for Seti and Einstein also. I don't understand why. I know Seti started an Email recruiting effort designed to get previous Seti users back. The rises seem to correspond with that, but I don't know why it would affect Rosetta and Einstein, unless the new (old) users attached to more than just Seti.
ID: 29527 · Rating: -1 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile jaxom1
Avatar

Send message
Joined: 5 Jun 06
Posts: 180
Credit: 1,586,889
RAC: 0
Message 29528 - Posted: 17 Oct 2006, 18:02:10 UTC

It's getting colder, so people can run their computers more and not have to worry about the heat they generate.

I live around Chicago. Maybe I should try to heat the apartment using nothing but computers this winter. LOL.. Or maybe not...

About the Seti e-mails. I can confirm they are indeed sending these, as I received one. I ran Seti before it was on boinc. See below for what they are sending out.



Dear Blah:

We notice that you registered with SETI@home on 30 December 2003, and that we haven't yet credited you with any completed work for SETI@home/BOINC. Your computing power is important to SETI@home. If you had problems installing or running the software, please visit the Questions and Answers area of our web site. You may find the solution to your problem there; if not, you can ask for help from project volunteers and staff.

We also encourage you to fill out the BOINC user survey. to give us feedback on how to make SETI@home easier to use.

Thanks!

-- The SETI@home team


If you'd like to not get any more email from us, please click here.

ID: 29528 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Michael G.R.

Send message
Joined: 11 Nov 05
Posts: 264
Credit: 11,247,510
RAC: 0
Message 29532 - Posted: 17 Oct 2006, 18:57:59 UTC - in response to Message 29528.  

Dear Blah:

We notice that you registered with SETI@home on 30 December 2003, and that we haven't yet credited you with any completed work for SETI@home/BOINC. Your computing power is important to SETI@home. If you had problems installing or running the software, please visit the Questions and Answers area of our web site. You may find the solution to your problem there; if not, you can ask for help from project volunteers and staff.


I got the same thing, except that (from memory) I signed up around 2000. Man, time flies!

ID: 29532 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
FluffyChicken
Avatar

Send message
Joined: 1 Nov 05
Posts: 1260
Credit: 369,635
RAC: 0
Message 29534 - Posted: 17 Oct 2006, 19:42:09 UTC

Maybe due to SIMAP going part time ?

Begining of the scholars season ?

.....
Team mauisun.org
ID: 29534 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Astro
Avatar

Send message
Joined: 2 Oct 05
Posts: 987
Credit: 500,253
RAC: 0
Message 29537 - Posted: 17 Oct 2006, 20:41:21 UTC - in response to Message 29534.  
Last modified: 17 Oct 2006, 20:44:40 UTC

Maybe due to SIMAP going part time ?

Begining of the scholars season ?

.....

BoincSimap hmmer 5.08 was not working well with AMD systems. I'm currently attached to Simap to collect data on the new hmmer 5.09 Wus. They are keeping me fully supplied. I too read the message on the main board at Simap, but for atleast the last few days, they seem to have work (especially the period of increased new users at other projects). I have a 1/2 day cache, so frequently get work.
ID: 29537 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Buffalo Bill
Avatar

Send message
Joined: 25 Mar 06
Posts: 71
Credit: 1,630,458
RAC: 0
Message 29550 - Posted: 18 Oct 2006, 2:41:27 UTC
Last modified: 18 Oct 2006, 2:42:12 UTC

I got an e-mail from Seti@Home back in March 2006. I had been with their project from June 1999 into 2000. So when I came back to have a look I discovered BOINC and joined Rosetta. In my case Rosetta gained from Seti's e-mail to me. I just got their new one too.
ID: 29550 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
David Baker
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 17 Sep 05
Posts: 705
Credit: 559,847
RAC: 0
Message 29555 - Posted: 18 Oct 2006, 3:59:26 UTC - in response to Message 29519.  

1092 new hosts in the past 24 hours, a bit more than that the day before. Nice little spike.

Maybe this has been discussed before, and if so, please just point me in the right direction, but I was wondering if the Rosetta team had some kind of threshold that they are aiming for before rolling out some new science.

f.ex., Do we need more than 40 teraflops of computational power before it makes sense to send out some more complex proteins? Do we need X number of teraflops before a new algorhythm is used or whatever?

In short: Are we waiting for a certain number of crunching power before implementing some things that are already ready to go?

Thanks,


Yes--I certainly hope this spike continues!

There isn't really a critical threshold, but the accuracy of both our structure prediction and design efforts are always limited by the total amount of computing power we (collectively) can bring to bear on the problem.
For example, we will soon be sending out on rosetta@home our calculations of the structure of the amyloid fiber responsible for Alzheimer's disease. This problem will take a particularly large amount of computing because it involves many protein chains.
The answer to your last question is that we send new work units with new approaches or problems out after we have completed testing on our in house computers and things look promising. also, as pointed out, we can't send out many of the design calculations, as they require too much memory (we have to consider all 20 amino acids at all positions in the protein in some cases which involves a lot of precomputation of large interaction tables)

ID: 29555 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Hoelder1in
Avatar

Send message
Joined: 30 Sep 05
Posts: 169
Credit: 3,915,947
RAC: 0
Message 29557 - Posted: 18 Oct 2006, 5:30:23 UTC - in response to Message 29555.  
Last modified: 18 Oct 2006, 5:31:47 UTC

also, as pointed out, we can't send out many of the design calculations, as they require too much memory (we have to consider all 20 amino acids at all positions in the protein in some cases which involves a lot of precomputation of large interaction tables)
I am sure there would be those who would install whatever memory is required to perform those large calculations on their computers... -H.
Team betterhumans.com - discuss and celebrate the future - hoelder1in.org
ID: 29557 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1830
Credit: 119,208,549
RAC: 2,278
Message 29565 - Posted: 18 Oct 2006, 9:17:39 UTC

We're moving off-topic, but maybe the project needs (or will soon need) splitting into categories for available RAM? I know there's an identifier for 1GB+ jobs. Rather than distinct groups, a continuous scale based on the requirements of the job would be better IMO. I know the amount of RAM available in the pool is increasing, but I assume Rosetta's requirements are expanding much more rapidly.

A job might need up to, say 700MB RAM in testing, so add a safety % to this and send this out to all computers that allow BOINC to use this much RAM in the preferences.

I guess the bigger jobs should start receiving more credit at some point. Maybe the job could track how much RAM it's using and these could be averaged just as the CPU benchmarks are currently. This would award credit based on the RAM used, which would in turn be dependent on the settings in the BOINC/Rosetta preferences regarding RAM/Virtual Mem useage.

cheers
Danny
ID: 29565 · Rating: -1 · rate: Rate + / Rate - Report as offensive    Reply Quote
Astro
Avatar

Send message
Joined: 2 Oct 05
Posts: 987
Credit: 500,253
RAC: 0
Message 29618 - Posted: 19 Oct 2006, 3:20:35 UTC

Update, Boincsimap has started cutting back on work. I received some 5.09hmmers' then shortly after I recieved some 5.10 wus. Then those dried up as well. I've allowed another project to feed these machines and hope they give atleast a few more so I have a decent number of samples.
ID: 29618 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Gen_X_Accord
Avatar

Send message
Joined: 5 Jun 06
Posts: 154
Credit: 279,018
RAC: 0
Message 30363 - Posted: 31 Oct 2006, 9:15:04 UTC - in response to Message 29528.  
Last modified: 31 Oct 2006, 9:17:50 UTC

It's getting colder, so people can run their computers more and not have to worry about the heat they generate.

I live around Chicago. Maybe I should try to heat the apartment using nothing but computers this winter. LOL.. Or maybe not...


You may laugh, but I'm doing it. I live in Nebraska, we get just as frozen.
As far as the heating with my computer, I'm trying it, but you have to understand that my apartment holds heat well in the winter, and stays very cool in the summer. I can shut off the heater or a/c when I am not at home, and the temp will be the same when I get home from work 11 hours later. I saw the temp in the apartment drop to only 65 once when I was out of town for 3 days. As you can guess, my utility bills are very low just based on that. So I'm trying to see if using the computer, running Rosetta all the time, is feasible for keeping my apartment warm. I'll have to wait for the bill and compare it to last year to see how it balances out with the electricity cost.(P.s. my apt. heater is gas and that is paid for in the rent, so I pay elect. only. I am comparing efficiency, as I am an efficiency nut.)
ID: 30363 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
zombie67 [MM]
Avatar

Send message
Joined: 11 Feb 06
Posts: 316
Credit: 6,621,003
RAC: 0
Message 30369 - Posted: 31 Oct 2006, 13:50:57 UTC - in response to Message 30363.  

You may laugh, but I'm doing it. I live in Nebraska, we get just as frozen.
As far as the heating with my computer, I'm trying it, but you have to understand that my apartment holds heat well in the winter, and stays very cool in the summer. I can shut off the heater or a/c when I am not at home, and the temp will be the same when I get home from work 11 hours later. I saw the temp in the apartment drop to only 65 once when I was out of town for 3 days. As you can guess, my utility bills are very low just based on that. So I'm trying to see if using the computer, running Rosetta all the time, is feasible for keeping my apartment warm. I'll have to wait for the bill and compare it to last year to see how it balances out with the electricity cost.(P.s. my apt. heater is gas and that is paid for in the rent, so I pay elect. only. I am comparing efficiency, as I am an efficiency nut.)


Try running only the fan (as needed) to redistribute the heat throughout the apartment.
Reno, NV
Team: SETI.USA
ID: 30369 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Jose

Send message
Joined: 28 Mar 06
Posts: 820
Credit: 48,297
RAC: 0
Message 30385 - Posted: 31 Oct 2006, 19:39:27 UTC - in response to Message 29557.  

also, as pointed out, we can't send out many of the design calculations, as they require too much memory (we have to consider all 20 amino acids at all positions in the protein in some cases which involves a lot of precomputation of large interaction tables)
I am sure there would be those who would install whatever memory is required to perform those large calculations on their computers... -H.



There were
This and no other is the root from which a Tyrant springs; when he first appears he is a protector.”
Plato
ID: 30385 · Rating: -1 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : 1000+ new hosts in past two days & question



©2024 University of Washington
https://www.bakerlab.org