Flakey AMD/ATI GPUs, including RX 5700 XT, Cross Validating, polluting the Database

Message boards : Number crunching : Flakey AMD/ATI GPUs, including RX 5700 XT, Cross Validating, polluting the Database
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 9 · 10 · 11 · 12 · 13 · 14 · 15 . . . 20 · Next

AuthorMessage
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026236 - Posted: 4 Jan 2020, 4:40:31 UTC

Another Xmas present has made a nuisance of itself here.

Tanis 10773581

Cheers.
ID: 2026236 · Report as offensive     Reply Quote
Profile Mr. Kevvy Crowdfunding Project Donor*Special Project $250 donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 3777
Credit: 1,114,826,392
RAC: 3,319
Canada
Message 2026271 - Posted: 4 Jan 2020, 12:09:10 UTC - in response to Message 2026236.  
Last modified: 4 Jan 2020, 12:35:50 UTC

Tanis, Richard and AMD Jesus (edit: and George Ko) pestered. stogdan and Richard Hartland also replied and indicated they are disabling their affected GPUs... yay!

The list is two pages back or more so here's a repost of it:

李溪伦 9302807
[AfZ]TomServo1 1483720
achimbln 138625
Alexandr Galushchenko 9609912
alffrommars 9024750
AMD Jesus 70887
antoi 10856207
aridhol 10288747
Arnab 10093567
Baldarov 9438496
Bigthor 480399
Borktron 10682716
Brandon 8198367
calendir 9663884

Camiron 7449359
Carl 914781
Christopher 9894096
CoffeeSloth 10266313
Crisu 7833612
dalex 10881818
Daniel Conrad Broom 8059986
Daniel frederikson 9813817
Daniel Penz 91581
Dank 49802
Derrek 219419
Doc_Jebus 10863878
dsharbour 10858679
Dzsozi 8002127
Earendil 146007
egon.sauter 494566
Eirikafh 10883218
Eric 9157146
eryndel 10878567
Esta 10624508
Foaming Mad Cow Industries 219464
fred 1935325
fredi 7913572
George Ko 639539
ghostbuster 564989
gunsnammo 137399
Haiko_N 9198068
HawkMedic 10838738
higemayuge 10790664
HMZ 9079227
Jeff 10639246
Jeffrey A. Smith 38247
Jerjes 1291426
JohnDoe 9166075
Jorge Barrera 9650295
Juraxell 10864786
Kekke 46817
knutella 9880098
lastsworder 10878688
lupaslupas 10002927
MadMikeDelta 8221690
Maulwurf 1516335
MaximusPrometheus 10240426
mgg 279419
mnelsonx 272885
Niflhuem 113140

No Name@Extraterrestrial Intelligence 8116
NYX.consulting 10503661
Oriah 9838773

Otosan 8547502
PantherJon 9801065
Peter Furlong 7965665
phoenix7477 10773411
Rafael 8249913
rame 10738
rAttmAniA 9002301
Recedham 954834
rgeens 10740140
Richard 8565733
Richard Hartland 9781177
Rocky 270621

Saint123 159425

Stephen Diem 36679
stogdan 10865456
StrayCat 177967

Strickland 34273
suhail ahmad 9878177
Swagstergo 10882690
T66 3336343
Tanis 10773581
toby 9442798
TomasFraus 8445239
Tomik 8972653
Trezy 10367889
Tristan 9778349
vleermuis 1295921
VMS Software Inc 45538

werewolf_007 10880222
xakei 10823091
Zac 100334866

Italicized names have replied and indicated they are disabling their affected GPUs.
Struck-through names are confirmed to no longer be producing these bad results, ie via disabling GPU computing.
ID: 2026271 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026273 - Posted: 4 Jan 2020, 12:29:33 UTC

It's been a while, but I've been ripped off again, https://setiathome.berkeley.edu/workunit.php?wuid=3820918292. :-(

And a new culprit, George Ko 639539.

Cheers.
ID: 2026273 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026360 - Posted: 5 Jan 2020, 1:00:02 UTC

Another new rig on the block. :-(

unbound 10885610

Cheers.
ID: 2026360 · Report as offensive     Reply Quote
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 2026366 - Posted: 5 Jan 2020, 1:31:05 UTC

new one: Hozer https://setiathome.berkeley.edu/show_host_detail.php?hostid=8878385

his prime card shows a R9 390, but his secondary card is a 5700XT.

also, for AMD cards, it appears that the level of double precision performance is the main/first factor for ranking the cards in BOINC.

also interesting, and I just noticed, that the seti apps only report half of the number of compute units on these new cards.
5700's report as 18 CU, when in reality they have 36
5700XTs report as 20 CU, when in reality they have 40
but the older cards seem to report correctly, eg. the R9 390 correctly reports 40 CUs
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 2026366 · Report as offensive     Reply Quote
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2026372 - Posted: 5 Jan 2020, 2:11:03 UTC

Richard and DA need to update coproc.cpp again to handle the Navi cards by adding code for cores_per_proc = parameter.

They had to change the code to properly handle the new Turing cards where the cores_per_proc value was half of Pascal.https://github.com/BOINC/boinc/pull/2707
Looks like they need to double the value for Navi.

https://github.com/BOINC/boinc/commit/20af3e90ce165c23edb700ecc21ec729dd36dff8
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2026372 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14656
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2026413 - Posted: 5 Jan 2020, 11:08:20 UTC - in response to Message 2026372.  
Last modified: 5 Jan 2020, 12:00:44 UTC

I'm a little hesitant about submitting this change, because I've never personally worked with AMD cards or with the developer of an AMD app. From what I understand, the process for deriving the card geometry from the API is rather different from the NV situation. Also, the only effect of the NV patch I submitted was the correct calculation of GFlops Peak, which helps in runtime estimation - it doesn't affect the actual calculations at all.

Having said that, I'm perfectly happy to submit a patch if we can work out exactly what is needed. I can speak with Ray Hinchliffe (author of SIV) to see how he handles those cards: we cross checked the NV results before submitting the last change. I'll also hunt through the Wiki reports, which are usually pretty comprehensive. There may even be manufacturer documentation... ? Anyone? ;-)

Edit - the ATI code (and comments) are at https://github.com/BOINC/boinc/blob/master/lib/coproc.cpp#L841

There don't seem to be any 'magic numbers' for cores_per_proc, like we have to use for NV - anybody really knowledgeable about ATI hardware willing to help?
ID: 2026413 · Report as offensive     Reply Quote
Profile Azmodes
Avatar

Send message
Joined: 28 Nov 16
Posts: 11
Credit: 6,317,066
RAC: 6
Austria
Message 2026421 - Posted: 5 Jan 2020, 12:41:51 UTC

If it helps, I cross-referenced my recent inconclusives with the list posted earlier and these three are candidates all running RX 5700 hosts that are not yet listed:

dcox https://setiathome.berkeley.edu/show_user.php?userid=10884993
iinkabob https://setiathome.berkeley.edu/show_user.php?userid=102908
Simgiov https://setiathome.berkeley.edu/show_user.php?userid=8796082
ID: 2026421 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026524 - Posted: 6 Jan 2020, 0:27:10 UTC

While Strickland gave me another mauling another new comer also joined in. :-(

killerepicprofurrygamer6969 10885981

Cheers.
ID: 2026524 · Report as offensive     Reply Quote
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 2026550 - Posted: 6 Jan 2020, 2:46:02 UTC

New one: BigDaddyDave - https://setiathome.berkeley.edu/show_host_detail.php?hostid=8874591
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 2026550 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026570 - Posted: 6 Jan 2020, 8:22:21 UTC

Well I've been ripped off again and it certainly doesn't look like Eric's fix is working. :-(

https://setiathome.berkeley.edu/workunit.php?wuid=3823542361

And another new culprit came with it, teargasm 10886461.

Cheers.
ID: 2026570 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026647 - Posted: 7 Jan 2020, 1:15:05 UTC

Mr. Kevvy could you check your list please as you have some unclosed tags in it I've just noticed as calendir 9663884 is still using their GPU.

Other than the usual culprits 4 new names turned up here.

AshlandPony 9004257
cprince1977 10886783
Illyria 10845292
jcr 3428

Cheers.
ID: 2026647 · Report as offensive     Reply Quote
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2026659 - Posted: 7 Jan 2020, 2:59:59 UTC - in response to Message 2026413.  

From what I understand, the process for deriving the card geometry from the API is rather different from the NV situation.

What isn't to say the API itself is incorrectly reporting the number of SM's on the newer cards. After neither Khronos nor AMD has fixed the drivers yet for this threads problem, I wouldn't bet either way.
It would be good to check with Ray and see if SIV reports the correct number of SM's or whether he had to modify/fudge SIV to fix what the API reports to jibe with what AMD says the architecture actually has.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2026659 · Report as offensive     Reply Quote
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22285
Credit: 416,307,556
RAC: 380
United Kingdom
Message 2026781 - Posted: 8 Jan 2020, 9:57:04 UTC

User Capizzi #10504781
computer #8860521
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 2026781 · Report as offensive     Reply Quote
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 2026811 - Posted: 8 Jan 2020, 13:43:34 UTC
Last modified: 8 Jan 2020, 13:43:46 UTC

firecrypt https://setiathome.berkeley.edu/show_host_detail.php?hostid=8869537
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 2026811 · Report as offensive     Reply Quote
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19158
Credit: 40,757,560
RAC: 67
United Kingdom
Message 2026843 - Posted: 8 Jan 2020, 17:37:41 UTC
Last modified: 8 Jan 2020, 17:39:10 UTC

ID: 2026843 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35197
Credit: 261,360,520
RAC: 489
Australia
Message 2026905 - Posted: 9 Jan 2020, 1:50:47 UTC

Just 1 new 1 for me today to report.

Adam Tadian 9295811

Cheers.
ID: 2026905 · Report as offensive     Reply Quote
Aravah

Send message
Joined: 16 Sep 19
Posts: 1
Credit: 357,627
RAC: 4
United Kingdom
Message 2027015 - Posted: 9 Jan 2020, 22:09:41 UTC

I am still keeping my AMD GPU firmly switched off despite regular messages from SETI@home app to fix this! :) I quote:

SETI@home: Notice from BOINC
Your settings do not allow fetching tasks for AMD/ATI GPU. To fix this, you can change Project Preferences on the project's web site.
Thu 09 Jan 2020 16:13:08 GMT

Perhaps it would help if SETI@home could send a different notification to those of us who have AMD GPUs until such time this OpenCL issue is resolved.

Ta
Aravah 10852549
ID: 2027015 · Report as offensive     Reply Quote
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 2027021 - Posted: 9 Jan 2020, 22:32:38 UTC - in response to Message 2027015.  

I am still keeping my AMD GPU firmly switched off despite regular messages from SETI@home app to fix this! :) I quote:
SETI@home: Notice from BOINC
Your settings do not allow fetching tasks for AMD/ATI GPU. To fix this, you can change Project Preferences on the project's web site.
Thu 09 Jan 2020 16:13:08 GMT
Perhaps it would help if SETI@home could send a different notification to those of us who have AMD GPUs until such time this OpenCL issue is resolved.
Ta
Aravah 10852549


. . You make a good point ...

Stephen

. .
ID: 2027021 · Report as offensive     Reply Quote
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13164
Credit: 1,160,866,277
RAC: 1,873
United States
Message 2027030 - Posted: 9 Jan 2020, 23:23:15 UTC

Just got apprised that AMD released new Adrenaline drivers for Windows 10 today. One of the fixed issues is with Seti@home result overflows.

Fixed result overflows that can be experienced with Radeon RX 5700 series when using SETI@Home.

https://www.amd.com/en/support/kb/release-notes/rn-rad-win-20-1-1
Need to have some of the known 5700XT users that have responded to PM's give the new drivers a test run.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 2027030 · Report as offensive     Reply Quote
Previous · 1 . . . 9 · 10 · 11 · 12 · 13 · 14 · 15 . . . 20 · Next

Message boards : Number crunching : Flakey AMD/ATI GPUs, including RX 5700 XT, Cross Validating, polluting the Database


 
©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.