rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7549
Credit: 44,264,218
RAC: 23,104
How long did the old app version take at you gpu?
For me, it looks like runtimes are quite identical between old and new app version.
There is no speed increase, the main change was only the output file.
/ 
Reply
Quote
The 3 took 34, 23, and 26 minutes. Can't remember what the old ones took on this card. Not completely out of line I'm sure and way less than the 45 minutes.
/ Reply
Quote
were new tasks released???
BOINC is saying task is 100% completed but the stderr file is still updating in the background? Is this supposed to happen??
Will continue testing to see what the final compute time is...
/ Reply
Quote
For me everything works fine, ~4 minutes compute time per WU.
Thanks!
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7549
Credit: 44,264,218
RAC: 23,104
mfakto apps v44 for intel arc on linux/windows updated
new
- support for new intel arc cards
/ Reply
Quote
Question would an OpenCL application for NVIDIA run more efficiently than the current Cuda 120 application ?
____________
In October 1969 I took an oath to support and defend the Constitution of the United States against all enemies, foreign and domestic;
There was no expiration date.
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7549
Credit: 44,264,218
RAC: 23,104
Question would an OpenCL application for NVIDIA run more efficiently than the current Cuda 120 application ?
cuda is much faster, the same stream vs. opencl from AMD in earlier days
/ Reply
Quote
Hello and good morning.
Because of the "news" with "should now work for RX6000, RX7000" i activated the TF v0.44 (opencl_ati_101) tasks for my AMD GPU ( a 7900 XTX on Adrenalin 24.12.1 on Windows 11) again, but have mainly errors... just 2 TF tasks (15177832 and 15234212 ) were successfull, the other 78 run in errors.
e.g.
<core_client_version>8.0.2</core_client_version>
<![CDATA[
<message>
Das Betriebssystem kann (null) nicht ausf�hren.
(0xc3) - exit code 195 (0xc3)</message>
<stderr_txt>
2025-01-10 20:31:14 (16932): wrapper (7.24.26018): starting
2025-01-10 20:31:14 (16932): wrapper: running mfakto.exe (-d 0)
2025-01-10 20:31:14 (16932): wrapper: created child process 18508
mfakto 0.16-beta.2-Win (64-bit build)
Runtime options
INI file mfakto.ini
Verbosity 1
SieveOnGPU yes
MoreClasses yes
GPUSievePrimes 81157
GPUSieveProcessSize 24 Kib
GPUSieveSize 96 Mib
FlushInterval 0
WorkFile worktodo.txt
ResultsFile results.txt
JSONResultsFile results.json.txt
LogFile mfakto.log
Checkpoints enabled
CheckpointDelay 60 s
Stages disabled
StopAfterFactor bitlevel
PrintMode compact
Logging disabled
V5UserID none
ComputerID none
TimeStampInResults no
VectorSize 2
GPUType AUTO
SmallExp no
UseBinfile mfakto_Kernels.elf
Compile-time options
Select device - 2025-01-10 20:31:23 (16932): mfakto.exe exited; CPU time 0.078125
2025-01-10 20:31:23 (16932): app exit status: 0xc0000374
2025-01-10 20:31:23 (16932): called boinc_finish(195)
</stderr_txt>
]]>
What can i do now to fix these errors?
THX Sputnik
/ Reply
Quote
Volunteer developer
Volunteer tester
Send message
Joined: 29 Nov 14
Posts: 88
Credit: 406,218,499
RAC: 1,715,944
This message:
<message>
Das Betriebssystem kann (null) nicht ausf�hren.
(0xc3) - exit code 195 (0xc3)</message>
is from your OS.
Can you restart your PC and look if this error disappears?
/ Reply
Quote
Moin & THX. The first TF task https://srbase.my-firewall.org/sr5/result.php?resultid=15350856 was done successfully after restarting the PC. It took "relatively long" 22min... 7900 XTX GPU took ~170W during crunching.
e.g.
...
no factor for M241356827 from 2^75 to 2^76 [mfakto 0.16-beta.2-Win cl_barrett32_76_gs_2]
tf(): total time spent: 22m 7.432s (2063.58 GHz-days / day)
ERROR: get_next_assignment(): no valid assignment found in "worktodo.txt"
2025-01-12 10:51:20 (21144): mfakto.exe exited; CPU time 10.093750
2025-01-12 10:51:20 (21144): called boinc_finish(0)
</stderr_txt>
]]>
Will crunch some more TF and looking for errors...
/ Reply
Quote
Had 51 more valid 0.44 TF tasks = my PC with AMD 7000series GPU is doing it correctly :D
THX
/ Reply
Quote
Programma
TF 0.44 (opencl_intel_gpu_intelarc)
Bestandsnaam
TF_75-76_183-241M_wu_76059
Status
Bezig met uploaden (opgeschort - op verzoek van gebruiker)
Ontvangen
13-1-2025 20:50:56
Rapporteren voor
18-1-2025 20:47:53
Bronnen
0.725 CPUs + 1 Intel GPU
Geschatte berekenings-grootte
80 GFLOPs
Processor tijd
00:03:49
Verstreken tijd
00:31:55
Uitvoerbaar bestand
wrapper_26012-v2_windows_x86_64.exe
/ Reply
Quote
Nummer 2
Programma
TF 0.44 (opencl_intel_gpu_intelarc)
Bestandsnaam
TF_75-76_183-241M_wu_84475
Status
Bezig met uploaden (opgeschort - op verzoek van gebruiker)
Ontvangen
13-1-2025 20:50:56
Rapporteren voor
18-1-2025 20:47:53
Bronnen
0.725 CPUs + 1 Intel GPU
Geschatte berekenings-grootte
80 GFLOPs
Processor tijd
00:26:26
Verstreken tijd
00:27:10
Uitvoerbaar bestand
wrapper_26012-v2_windows_x86_64.exe
180W
/ Reply
Quote
TF 0.44 downloads at me as a 1 sec. job, but takes hours...
Subsequently, it downloads about 60 of them -claiming it to take 1 minute-
and starts to do them at high priority, as there are more projects running on the same machine. Before 20 have finished, the rest gets aborted due to not having been started in time.
/ Reply
Quote
Before 20 have finished, the rest gets aborted due to not having been started in time.
This also happened to me when I first started processing the new work all my work completed in time.
As long as you are/completing returning successful work I would ask for work again as your machine theoretically shouldn't ask for as many meaning you should be able to get through them before the deadline has passed.
I am unable to look at your hardware as your hosts are "hidden"
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7549
Credit: 44,264,218
RAC: 23,104
The main problem with estimated runtime and too many tasks are after installing or updating new apps. The server didnt know them and do a reset of client datas. After some results are back it should be back to normal.
/ Reply
Quote
Completely from a curiosity point may I ask what M range the test work was from?
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7549
Credit: 44,264,218
RAC: 23,104
Completely from a curiosity point may I ask what M range the test work was from?
range 67,68 was the test, 75,76 is currently running
/ Reply
Quote