GBT ('guppi') .vlar tasks will be send to GPUs, what you think about this?

Message boards : Number crunching : GBT ('guppi') .vlar tasks will be send to GPUs, what you think about this?
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 7 · 8 · 9 · 10

AuthorMessage
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13370
Credit: 208,696,464
RAC: 304
Australia
Message 1788307 - Posted: 18 May 2016, 6:02:54 UTC - in response to Message 1788192.  

There is another possibilities like task re-scheduling that will do the same w/o any harm to project itself.

I thought task rescheduling screwed up granted credit even more than Credit New does?
Grant
Darwin NT
ID: 1788307 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 21016
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1788310 - Posted: 18 May 2016, 6:20:06 UTC

Re-scheduling does screw the credit granted for a task, and also degrades the base level of credit for all tasks. This could well explain some of the drop in credit per hour we are all seeing.
NEVER, EVER advise anyone to reschedule between CPU and GPU, or CPU families (Intel - Nvidia), further, if you are re-scheudling STOP DOING SO IMMEDIATELY.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1788310 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1788314 - Posted: 18 May 2016, 6:29:37 UTC - in response to Message 1788310.  

Thanks Rob, you saved me a bunch of reading / configuring :)
ID: 1788314 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1788317 - Posted: 18 May 2016, 6:45:18 UTC - in response to Message 1788314.  

It makes me wonder now ... Is this how people are getting thousands of tasks on a device?
ID: 1788317 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6324
Credit: 106,370,077
RAC: 121
Russia
Message 1788320 - Posted: 18 May 2016, 6:56:44 UTC - in response to Message 1788307.  

There is another possibilities like task re-scheduling that will do the same w/o any harm to project itself.

I thought task rescheduling screwed up granted credit even more than Credit New does?

Consider and compare 2 options:
1) to abort task
2) to re-schedule task.

1 - if 10 peoples do this task will lost. Completely. Any data it contains will be lost.
2- task processed, project science not harmed at all. Credits _possibly_ little changed. I saw no definitive prove that credit will be screwed, only theories. And definitely not in such degree to worry about, consider overflows get full credit and x2 credit mis-granting common.

What one who interested in project results, not in competition for imaginary toaster, will chose?
ID: 1788320 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13370
Credit: 208,696,464
RAC: 304
Australia
Message 1788322 - Posted: 18 May 2016, 7:04:24 UTC - in response to Message 1788320.  
Last modified: 18 May 2016, 7:05:12 UTC

1 - if 10 peoples do this task will lost. Completely. Any data it contains will be lost.

So no matter how often people say Credits aren't important, this indicates they are. And as it runs the risk of impacting on the science being done it would bolster the argument that something needs to be done to fix Credit New.
IMHO.


What one who interested in project results, not in competition for imaginary toaster, will chose?

Just to let the WUs run on the GPU.
Grant
Darwin NT
ID: 1788322 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6324
Credit: 106,370,077
RAC: 121
Russia
Message 1788326 - Posted: 18 May 2016, 7:12:34 UTC - in response to Message 1788322.  
Last modified: 18 May 2016, 7:13:00 UTC

1 - if 10 peoples do this task will lost. Completely. Any data it contains will be lost.

So no matter how often people say Credits aren't important, this indicates they are. And as it runs the risk of impacting on the science being done it would bolster the argument that something needs to be done to fix Credit New.
IMHO.

Absolutely no objections on this point. I would little re-formulate it as "to fix credit granting scheme for SETI project". Actually, this could be more really doable than fixing CreditScrew itself...


What one who interested in project results, not in competition for imaginary toaster, will chose?

Just to let the WUs run on the GPU.

Again, that third choice will be best one indeed. In long run constructive suggestions and tuning/run experience will improve GPU performance even for this type of work. But some wanted to exclude it from consideration.

There is good suggestion in this thread about server scheduler priorities - worth to present it to Eric for possible implementation.
ID: 1788326 · Report as offensive
Kevin Olley

Send message
Joined: 3 Aug 99
Posts: 906
Credit: 261,085,289
RAC: 572
United Kingdom
Message 1788932 - Posted: 20 May 2016, 7:43:32 UTC

Time for a bit of selective cheating I thinks:-)

Just so that more WU's can be processed.

Don't worry, all WU's will be done.

I am not aborting or re-scheduling any WU's.
Kevin


ID: 1788932 · Report as offensive
Chris Oliver Project Donor
Avatar

Send message
Joined: 4 Jul 99
Posts: 72
Credit: 134,288,250
RAC: 15
United Kingdom
Message 1788944 - Posted: 20 May 2016, 8:34:32 UTC - in response to Message 1788320.  
Last modified: 20 May 2016, 8:40:28 UTC

Is anybody else having an issue with false positives from Windows10 Defender software when it scans the SOG application ?
ID: 1788944 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1788947 - Posted: 20 May 2016, 8:45:07 UTC - in response to Message 1788944.  

What one who interested in project results, not in competition for imaginary toaster, will chose?


It depends on whether such person would consider the results as the only part of the experiment, or if they consider the pain of change as part of the science too.

IOW: probably making the choices en-masse that feel right as individuals, add together to guide where things will head next. At some point the imaginary toaster/credits will be important, but for now it's more about understanding who's affected most.

Nash's Equilibrium (game theory) would suggest the most helpful situation would be for everyone to run as horribly as possible, so as to expose all the weaknesses. For some 'horribly as possible' would mean aborting, for others rescheduling, for others crunching CPU only. Probably since there aren't known 'best options' yet, I'd say we're likely to end up with a good outcome.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1788947 · Report as offensive
Profile William
Volunteer tester
Avatar

Send message
Joined: 14 Feb 13
Posts: 2037
Credit: 17,689,662
RAC: 0
Message 1788949 - Posted: 20 May 2016, 8:53:11 UTC - in response to Message 1788947.  

What one who interested in project results, not in competition for imaginary toaster, will chose?


It depends on whether such person would consider the results as the only part of the experiment, or if they consider the pain of change as part of the science too.

IOW: probably making the choices en-masse that feel right as individuals, add together to guide where things will head next. At some point the imaginary toaster/credits will be important, but for now it's more about understanding who's affected most.

Nash's Equilibrium (game theory) would suggest the most helpful situation would be for everyone to run as horribly as possible, so as to expose all the weaknesses. For some 'horribly as possible' would mean aborting, for others rescheduling, for others crunching CPU only. Probably since there aren't known 'best options' yet, I'd say we're likely to end up with a good outcome.

Ever the optimist, eh?
OTOH, maybe it's that bad, it can only get better...
A person who won't read has no advantage over one who can't read. (Mark Twain)
ID: 1788949 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1788950 - Posted: 20 May 2016, 8:57:34 UTC - in response to Message 1788949.  

Ever the optimist, eh?
OTOH, maybe it's that bad, it can only get better...


Also, 'Any port in a [sh#t]storm', lol
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1788950 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51464
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1788980 - Posted: 20 May 2016, 13:36:51 UTC - in response to Message 1788950.  

Ever the optimist, eh?
OTOH, maybe it's that bad, it can only get better...


Also, 'Any port in a [sh#t]storm', lol

The kitties are waiting for it to get better...
My output is less than half what it was a few weeks ago.
RAC still in decline.

Meowsigh.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1788980 · Report as offensive
J. Mileski
Volunteer tester
Avatar

Send message
Joined: 9 Jun 02
Posts: 632
Credit: 172,116,532
RAC: 572
United States
Message 1789072 - Posted: 20 May 2016, 21:19:05 UTC - in response to Message 1788944.  

Gave me a Trojan alert.
"Trojan:Win32/Varpes.M!cl"
file:C:\ProgramData\Microsoft\Windows Defender\LocalCopy\{834C27FD-CB6E-4487-937B-5D53DBC468F4}-mb8_win_x86_sse3_opencl_nv_r3430_sog.7z

I had downloaded OpenCL NV MultiBeam v8 SoG edition for Windows I think from Raistmer's link but I havn't had time to do anything with it yet.

Just letting you know, that M$ is always watching.

Is anybody else having an issue with false positives from Windows10 Defender software when it scans the SOG application ?


http://setiathome.berkeley.edu/forum_thread.php?id=79642
ID: 1789072 · 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 1789101 - Posted: 21 May 2016, 0:52:31 UTC

While checking the one machine I have going I notice Sog and Sah units .

Not only do they take longer but you don't get any extra credit for it .

FIX THE CREDIT SCREW it is getting out of hand or fire the rick that refuses to seriously look into it

The credits are the only reward there is and this B/S about it only being about the science is B/S and we all know it

I suggest you follow Einstein the longer it takes to do a unit the higher the reward

Cobblestones is irrelevant theses days with Multi cored CPU's and extremely fast GPU's

CPU doing AP's nearly 2 days doing same on GPU 2 hrs 20 mins but SAME CREDIT that's total B/S

Fix the credit system and lose that f-ing EGO you have and do what the USERS are asking .

No names needed as the person knows whom I'm talking about
ID: 1789101 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13370
Credit: 208,696,464
RAC: 304
Australia
Message 1789113 - Posted: 21 May 2016, 1:45:12 UTC - in response to Message 1789101.  
Last modified: 21 May 2016, 1:46:08 UTC

CPU doing AP's nearly 2 days doing same on GPU 2 hrs 20 mins but SAME CREDIT that's total B/S

No, that's the way things should be- getting the same amount of credit for doing the same amount of work.
If you're able to more WUs in a shorter period of time, then you will get more credit per day than those that can't do as many in the same period.
Grant
Darwin NT
ID: 1789113 · 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 1789128 - Posted: 21 May 2016, 2:20:55 UTC - in response to Message 1789113.  

CPU doing AP's nearly 2 days doing same on GPU 2 hrs 20 mins but SAME CREDIT that's total B/S

No, that's the way things should be- getting the same amount of credit for doing the same amount of work.
If you're able to more WUs in a shorter period of time, then you will get more credit per day than those that can't do as many in the same period.


Yes you are correct Grant wrong example to use .

Theses new SoG and Sah files thou are a joke the GPU seems to be only working 1/3 the time if that and times are over 2 hrs !!!!!

I've had to stop doing 3 and go to 2 as the times where soming like 6 hrs doing 3 threads compared to 1 hr 5 mins doing Cuda with 3 threads

(opps I realized I haven't done AP's on the CPU for long time forgot how long they took to do )
ID: 1789128 · Report as offensive
Profile tullio
Volunteer tester

Send message
Joined: 9 Apr 04
Posts: 8646
Credit: 2,930,782
RAC: 1
Italy
Message 1789175 - Posted: 21 May 2016, 6:03:16 UTC - in response to Message 1789101.  

On Einstein@home I get 4400 credits for a GPU task and 2000 credits for a CPU task. The CPU task takes much longer.
Tullio
ID: 1789175 · Report as offensive
Previous · 1 . . . 7 · 8 · 9 · 10

Message boards : Number crunching : GBT ('guppi') .vlar tasks will be send to GPUs, what you think about this?


 
©2022 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.