Mismatch in time elapsed values?

Message boards : Number crunching : Mismatch in time elapsed values?
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile SilentArceus
Volunteer tester

Send message
Joined: 4 Jan 11
Posts: 12
Credit: 17,461
RAC: 0
Message 1581829 - Posted: 4 Oct 2014, 14:37:31 UTC
Last modified: 4 Oct 2014, 14:38:11 UTC

I'm running my second atropulse wu right now and it's showing a gross mismatch between the (correct) value shown by boinc manager and (wrong) value shown by the astropulse graphics program. As of writing, BOINC is showing a time elapsed of about 81 hours and astropulse shows 2 days 22 hours (=nearly 70 hours). Now before you say anything lemme tell you BOINC gets nearly 100% of both processors almost everytime and ap is the only wu I'm running for hours (dual core machine).

Will the credits be computed based on the BOINC manager time or the ap graphics time?

(This is important for me to know as the ap credits would be much less than my fair share)
ID: 1581829 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34253
Credit: 79,922,639
RAC: 80
Germany
Message 1581840 - Posted: 4 Oct 2014, 14:52:56 UTC

The time has nothing to do with the credits you`ll get granted.


With each crime and every kindness we birth our future.
ID: 1581840 · Report as offensive
Darth Beaver Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 20 Aug 99
Posts: 6728
Credit: 21,443,075
RAC: 3
Australia
Message 1581851 - Posted: 4 Oct 2014, 15:18:41 UTC - in response to Message 1581829.  

I just looked at your machine you must have only just started doing AP's so until you finish and validate 10 units the elasped times will be out as the servers don't know yet what your machine can do .

It will correct it's self after you have done the 10 units .

Also there is a small problem with one of the server , the validation server so Astropulse units are not being validated at the moment . There know about it but it wont be fixed now till at least Monday

You will get the credits for them when they fix the validation servers

You can go back to only doing MB till they fix it but you may not get the credits if you keep going them till Tuesday or Wednesday
ID: 1581851 · Report as offensive
Profile SilentArceus
Volunteer tester

Send message
Joined: 4 Jan 11
Posts: 12
Credit: 17,461
RAC: 0
Message 1581854 - Posted: 4 Oct 2014, 15:25:04 UTC - in response to Message 1581851.  

You can go back to only doing MB till they fix it but you may not get the credits if you keep going them till Tuesday or Wednesday


I dont do any other SETI work when I'm doing an astropulse. I didnt exactly get the point here. Do you mean I shouldnt take/upload any work during The Holy Day of Maintenance?
ID: 1581854 · Report as offensive
Darth Beaver Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 20 Aug 99
Posts: 6728
Credit: 21,443,075
RAC: 3
Australia
Message 1581859 - Posted: 4 Oct 2014, 15:35:32 UTC - in response to Message 1581854.  

I'm saying if you keep doing AP's then you may not get the credits till after Tuesdays Maintenance day .

So if credit is important to you as it is with many others don't do AP's till the problem is fixed either Monday or after the Maintenance day

And i am saying you elasped times will not be correct till you do 10 units and they are VALIDATED

The problem with the servers is validation of AP's at the moment
ID: 1581859 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1581940 - Posted: 4 Oct 2014, 18:11:45 UTC

IIRC time shown on the graphics is CPU time, always somewhat less than the Elapsed time shown by BOINC, but having Elapsed 15 or 16 percent larger than CPU is unusual unless the system is being used for other things. One thing which might account for the difference is looking at the graphics if whatever is driving the display doesn't have effective OpenGL support. In that case the CPU has to do OpenGL processing, which slows down crunching. I doubt that's the issue on a Pentium 4 system, but it's possible.

The effective claimed credit does use the measured elapsed time (the server estimates are not involved), but it is combined with an even worse guess at the peak FLOPS of the CPU, then scaled in a way which is intended to normalize the credits. But since your host doesn't yet have enough tasks completed, the credit will mainly be determined by your Wingmate's effective claimed credit anyhow.

Note that more Elapsed time means a higher effective credit claim for an individual task, though that then feeds into an average which should slightly reduce credits for later tasks.
                                                                   Joe
ID: 1581940 · Report as offensive
Profile SilentArceus
Volunteer tester

Send message
Joined: 4 Jan 11
Posts: 12
Credit: 17,461
RAC: 0
Message 1582152 - Posted: 5 Oct 2014, 7:36:22 UTC - in response to Message 1581940.  
Last modified: 5 Oct 2014, 7:38:32 UTC

um...
ID: 1582152 · Report as offensive
Profile SilentArceus
Volunteer tester

Send message
Joined: 4 Jan 11
Posts: 12
Credit: 17,461
RAC: 0
Message 1582153 - Posted: 5 Oct 2014, 7:36:40 UTC - in response to Message 1581940.  

I've disabled the screensaver since when I first started SETI. The non-graphics program certainly doesent have any use for OpenGL does it?
ID: 1582153 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1582161 - Posted: 5 Oct 2014, 7:47:55 UTC - in response to Message 1582153.  

I've disabled the screensaver since when I first started SETI. The non-graphics program certainly doesent have any use for OpenGL does it?

Most people crunching probably no longer use the screen saver as it actually slows down crunching.
Grant
Darwin NT
ID: 1582161 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1582230 - Posted: 5 Oct 2014, 14:36:34 UTC - in response to Message 1582153.  

I've disabled the screensaver since when I first started SETI. The non-graphics program certainly doesent have any use for OpenGL does it?

Yes, if you only show the graphics temporarily there should be little effect even if OpenGL must be done on CPU.
                                                                   Joe
ID: 1582230 · Report as offensive
Darth Beaver Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 20 Aug 99
Posts: 6728
Credit: 21,443,075
RAC: 3
Australia
Message 1582377 - Posted: 6 Oct 2014, 3:23:51 UTC - in response to Message 1582153.  

Lord Assbutt the Fourth

while looking at your machine i notice it says you have 2 processor on a P4 chip , I'm think to have that you must have Hypo Threading turned on . If you wish to speed things up Go into your BIOS and turn it off . That's not good running Seti . Or having it on with just a P4 .It effectively slows the machine down . I have had a similar set up and i turn it off because it slowed things down way to much it runs a half speed when you have H/T on .basicaly in your case each processor is running at 1.4 gig not 2.8 It's ok for normal computer use but not Seti unless it is a more advanced chip with more cores than only 1 which is all a P4 has .
It looks and sounds good to see 2 processors but it's all smoke and mirrors when it comes to a P4 chip your better running without Hypo Threading off to do seti unless you get a better chip

Core 2 duo
Core 2 Quad
I3
I5
I7
and not all of these chips will have Hypo Threading either some do some don't
ID: 1582377 · Report as offensive
Profile BilBg
Volunteer tester
Avatar

Send message
Joined: 27 May 07
Posts: 3720
Credit: 9,385,827
RAC: 0
Bulgaria
Message 1582408 - Posted: 6 Oct 2014, 5:46:02 UTC - in response to Message 1581829.  

As of writing, BOINC is showing a time elapsed of about 81 hours and astropulse (graphics) shows 2 days 22 hours (=nearly 70 hours)

If you restart BOINC they should be on par

This is only cosmetic - BOINC will report the correct Run time to the servers
 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 
ID: 1582408 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1582574 - Posted: 6 Oct 2014, 15:48:34 UTC - in response to Message 1582377.  

...
I'm think to have that you must have Hypo Threading turned on . If you wish to speed things up Go into your BIOS and turn it off . That's not good running Seti . Or having it on with just a P4 .
...

OTOH, many users have reported a 25% or so productivity increase when using Intel Hyperthreading. That is, individual tasks take less than twice as long when running two per physical core.

It probably depends on various factors related to cache size, cache speed, main memory speed, etc. and the sensible way to determine whether to have it on or not is to actually see which way works best on the system. IOW, try each for long enough to see how many tasks get done per day.
                                                                   Joe
ID: 1582574 · Report as offensive
Profile ivan
Volunteer tester
Avatar

Send message
Joined: 5 Mar 01
Posts: 783
Credit: 348,560,338
RAC: 223
United Kingdom
Message 1582727 - Posted: 6 Oct 2014, 21:11:53 UTC - in response to Message 1582574.  

...
I'm think to have that you must have Hypo Threading turned on . If you wish to speed things up Go into your BIOS and turn it off . That's not good running Seti . Or having it on with just a P4 .
...

OTOH, many users have reported a 25% or so productivity increase when using Intel Hyperthreading. That is, individual tasks take less than twice as long when running two per physical core.

It probably depends on various factors related to cache size, cache speed, main memory speed, etc. and the sensible way to determine whether to have it on or not is to actually see which way works best on the system. IOW, try each for long enough to see how many tasks get done per day.
                                                                   Joe

Well, I've been running two identical machines 7371751 and 7371070 for more than a month now. Admittedly the one using hyperthreading has been crunching for one day longer than the other, but it has a 14-15% advantage over the non-HT machine in both total credit and RAC. Unfortunately the non-HT machine will have to be shut off tomorrow while an engineer investigates a small anomaly in its RAID controller, so it will slip behind a bit more. When we start using the HT machine in earnest, I'll have to turn off HT as we will then want the fastest individual processes, not the highest throughput.
ID: 1582727 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1582731 - Posted: 6 Oct 2014, 21:24:31 UTC - in response to Message 1582727.  

...
I'm think to have that you must have Hypo Threading turned on . If you wish to speed things up Go into your BIOS and turn it off . That's not good running Seti . Or having it on with just a P4 .
...

OTOH, many users have reported a 25% or so productivity increase when using Intel Hyperthreading. That is, individual tasks take less than twice as long when running two per physical core.

It probably depends on various factors related to cache size, cache speed, main memory speed, etc. and the sensible way to determine whether to have it on or not is to actually see which way works best on the system. IOW, try each for long enough to see how many tasks get done per day.
                                                                   Joe

Well, I've been running two identical machines 7371751 and 7371070 for more than a month now. Admittedly the one using hyperthreading has been crunching for one day longer than the other, but it has a 14-15% advantage over the non-HT machine in both total credit and RAC. Unfortunately the non-HT machine will have to be shut off tomorrow while an engineer investigates a small anomaly in its RAID controller, so it will slip behind a bit more. When we start using the HT machine in earnest, I'll have to turn off HT as we will then want the fastest individual processes, not the highest throughput.

For each processor I have tested HT or no HT from the P4 Northwood up to i7 Lynnfield I have seen gains with HT on. Most of my testing has been for machines that are CPU only. Some users with several GPUs in one system have reported gains with HT off.
Also in my testing I found leaving HT enabled & running half the threads produced slightly higher output than HT disabled.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1582731 · Report as offensive
Darth Beaver Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 20 Aug 99
Posts: 6728
Credit: 21,443,075
RAC: 3
Australia
Message 1582846 - Posted: 7 Oct 2014, 2:35:42 UTC - in response to Message 1582731.  

Hal have you done the same thing running just AP's and not MB ?
Did you get the same results ?
When i did it the times blow out so much and hung the machine when ever i try'd to do some thing it just was not worth it
ID: 1582846 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1582885 - Posted: 7 Oct 2014, 4:26:43 UTC - in response to Message 1582846.  

Hal have you done the same thing running just AP's and not MB ?
Did you get the same results ?
When i did it the times blow out so much and hung the machine when ever i try'd to do some thing it just was not worth it

I tested app configurations of:
only MB
only AP
half MB half AP
I tested system configurations of:
HT On, running on all threads
HT On, running on half the threads
HT Off, running on all cores
HT Off, running on half the cores
The only system I have found that didn't give maximum results running all threads with HT was is my dual socket 24 core server. Which I limit to running only 20 AP at a time. 24 seems to be to many for the amount of memory bandwidth the system has. In which I tried NUMA enabled & NUMA disabled.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1582885 · Report as offensive
Profile petri33
Volunteer tester

Send message
Joined: 6 Jun 02
Posts: 1668
Credit: 623,086,772
RAC: 156
Finland
Message 1583675 - Posted: 8 Oct 2014, 21:47:44 UTC - in response to Message 1582731.  


Also in my testing I found leaving HT enabled & running half the threads produced slightly higher output than HT disabled.
<-- I second to that.

My quess is that the processor is utilized to the max processes using real FPU's and then the additional fake CPU cores are used for whatever other (integer/logic) processing is needed.
To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones
ID: 1583675 · Report as offensive
Profile SilentArceus
Volunteer tester

Send message
Joined: 4 Jan 11
Posts: 12
Credit: 17,461
RAC: 0
Message 1584713 - Posted: 10 Oct 2014, 15:54:01 UTC

Thanks for all your suggestions. I've sent the results of that wu but it's pending validation so I'm yet to know it gets points on the basis of which time.

I might disable hyperthreading for some days and see it's effect some day.
ID: 1584713 · Report as offensive

Message boards : Number crunching : Mismatch in time elapsed values?


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.