Advanced search

Message boards : Graphics cards (GPUs) : 2800ti problem?

Author Message
Andy_Taximan
Send message
Joined: 19 Sep 14
Posts: 4
Credit: 20,278,250
RAC: 0
Level
Pro
Scientific publications
watwatwat
Message 52333 - Posted: 23 Jul 2019 | 19:53:16 UTC

Stderr output
<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
(unknown error) - exit code -59 (0xffffffc5)</message>
<stderr_txt>
# GPU [GeForce RTX 2080 Ti] Platform [Windows] Rev [3212] VERSION [80]
# SWAN Device 0 :
# Name : GeForce RTX 2080 Ti
# ECC : Disabled
# Global mem : 11264MB
# Capability : 7.5
# PCI ID : 0000:42:00.0
# Device clock : 1650MHz
# Memory clock : 7000MHz
# Memory width : 352bit
# Driver version : r431_31 : 43136
#SWAN: FATAL: cannot find image for module [.nonbonded.cu.] for device version 750

</stderr_txt>
]]>

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,572,627,986
RAC: 0
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 52334 - Posted: 23 Jul 2019 | 19:54:36 UTC

Not supported yet

Andy_Taximan
Send message
Joined: 19 Sep 14
Posts: 4
Credit: 20,278,250
RAC: 0
Level
Pro
Scientific publications
watwatwat
Message 52335 - Posted: 23 Jul 2019 | 20:01:24 UTC - in response to Message 52334.

That would explain it, such a new card out eh

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,572,627,986
RAC: 0
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 52337 - Posted: 23 Jul 2019 | 20:33:40 UTC - in response to Message 52335.

They are working on new software right now, the Linux users are all down too because of an expired license, so I would imagine it wouldn't be to long (that's a big portion of crunchers).

Jozef J
Send message
Joined: 7 Jun 12
Posts: 112
Credit: 1,118,845,172
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 52339 - Posted: 23 Jul 2019 | 21:48:32 UTC
Last modified: 23 Jul 2019 | 21:49:37 UTC

2080ti relase September 27, 2018
[url]https://en.wikipedia.org/wiki/GeForce_20_series
"new" ))
on milkyway boinc project is still best price/day rac 2012 year, old 7970 ati/amd gf card
detto einstein home, only amd card are reliable make some progress on project
nvidia hw tech is from another space ,humans dont know use ))))
[/url]

Keith Myers
Send message
Joined: 13 Dec 17
Posts: 1288
Credit: 5,097,631,959
RAC: 8,944,362
Level
Tyr
Scientific publications
watwatwatwatwat
Message 52340 - Posted: 24 Jul 2019 | 5:15:41 UTC

We Linux users are being held back by the Windows users. Our Turing compatible science application has been ready to use for months. We don't have access to it because we assume the scientists won't release it till they have figured out how to make and release an equivalent Windows application so there will be parity.

I'm getting tired of the wait. They should just release the Linux application so that 38% of the GPUGrid.net user base can start producing work again. Let the Windows users spin their wheels for a change for a couple of months like we have had to do since February.

mmonnin
Send message
Joined: 2 Jul 16
Posts: 332
Credit: 4,006,821,065
RAC: 12,216,791
Level
Arg
Scientific publications
watwatwatwatwat
Message 52343 - Posted: 24 Jul 2019 | 10:15:16 UTC

They should have just renewed the license and worked on this in the background.

Keith Myers
Send message
Joined: 13 Dec 17
Posts: 1288
Credit: 5,097,631,959
RAC: 8,944,362
Level
Tyr
Scientific publications
watwatwatwatwat
Message 52344 - Posted: 24 Jul 2019 | 20:03:54 UTC - in response to Message 52343.

They should have just renewed the license and worked on this in the background.

+ 1

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2343
Credit: 16,201,255,749
RAC: 1,038
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 52345 - Posted: 24 Jul 2019 | 22:40:49 UTC - in response to Message 52340.

We Linux users are being held back by the Windows users. Our Turing compatible science application has been ready to use for months. We don't have access to it because we assume the scientists won't release it till they have figured out how to make and release an equivalent Windows application so there will be parity.

I'm getting tired of the wait. They should just release the Linux application so that 38% of the GPUGrid.net user base can start producing work again. Let the Windows users spin their wheels for a change for a couple of months like we have had to do since February.
While I would like to use my Turing cards crunching for GPUGrid instead of SETI@home (under Linux) just like you, I think this is not as easily done from the project's side as you imply: The new app needs a wrapper, while the old app doesn't. That means the work generated for (and done by) the old app (most probably) won't work with the new app. So not just a different queue is required for the new app, but it should be managed separately which requires manpower which this project is lacking of.

Keith Myers
Send message
Joined: 13 Dec 17
Posts: 1288
Credit: 5,097,631,959
RAC: 8,944,362
Level
Tyr
Scientific publications
watwatwatwatwat
Message 52346 - Posted: 24 Jul 2019 | 23:14:58 UTC - in response to Message 52345.

A wrapper is just that. A wrapper around the science app. The wrapper does not and should not affect the workings of the science app. It is the same science app we have used in the past only rewritten to recognize the Turing cards. The result files for the old expired license Linux app should be exactly the same as the new wrapped app.

One of the reasons for going with the wrapper app is to get away from the micro-management of the science app for things like having to renew the license every year. It is also easier to deploy across platforms according to the wrapper documentation.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2343
Credit: 16,201,255,749
RAC: 1,038
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 52347 - Posted: 24 Jul 2019 | 23:49:07 UTC - in response to Message 52346.

It is the same science app we have used in the past only rewritten to recognize the Turing cards.
No, it's a completely different app. See this post.

Andy_Taximan
Send message
Joined: 19 Sep 14
Posts: 4
Credit: 20,278,250
RAC: 0
Level
Pro
Scientific publications
watwatwat
Message 52348 - Posted: 25 Jul 2019 | 1:08:29 UTC
Last modified: 25 Jul 2019 | 1:09:08 UTC

Not a good time to be picked as Formula Boinc Sprint then, no AMD and not even any cpu tasks :(

Keith Myers
Send message
Joined: 13 Dec 17
Posts: 1288
Credit: 5,097,631,959
RAC: 8,944,362
Level
Tyr
Scientific publications
watwatwatwatwat
Message 52349 - Posted: 25 Jul 2019 | 13:45:22 UTC - in response to Message 52347.

It is the same science app we have used in the past only rewritten to recognize the Turing cards.
No, it's a completely different app. See this post.

OK, conceded. I thought the previous app was acemd3 also.

Post to thread

Message boards : Graphics cards (GPUs) : 2800ti problem?

//