I tried both ... (separate directories)
They both ended normally with a results.txt and boinc_finish_called.
What info do you need?
/ 
Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
I tried both ... (separate directories)
They both ended normally with a results.txt and boinc_finish_called.
What info do you need?
no error in results.txt?
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
If someone has some issues and got this code:
195 (0xc3) EXIT_CHILD_FAILED
pls test it in standalone:
https://srbase.my-firewall.org/sr5/forum_thread.php?id=1449&postid=7402
Iam trying to figure out whats wrong.
The last successful result came from a RTX3060.
/ Reply
Quote
Just tried the V4 form the project directory with its wrapper (26012-v2) in a testdirectory.
Just runs OK like the 2 others.
Allowed new tasks ... all erroring out ...
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
Just tried the V4 form the project directory with its wrapper (26012-v2) in a testdirectory.
Just runs OK like the 2 others.
Allowed new tasks ... all erroring out ...
yeah thats strange, maybe the boinc client doesnt like it.
/ Reply
Quote
I just read somewhere that 3 digit error codes are BOINC (cliënt) errors, not the application...
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
Just tried the V4 form the project directory with its wrapper (26012-v2) in a testdirectory.
Just runs OK like the 2 others.
Allowed new tasks ... all erroring out ...
Are you running an antivirus program that can block the wrapper in the BOINC folder? Just checked the 2 win cuda apps and all looks the same.
That is the only one who can run the app:
[BOINC|7.16.11][CUDA|GeForce RTX 3060 Ti|1|4095MB|46172|102]
Update:
I will try to setup the recompiled app and add the cudart64_110.dll in the zip
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
v18 is up with included cudart64_110.dll
/ Reply
Quote
v18 is up with included cudart64_110.dll
Up and running !!!
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
v18 is up with included cudart64_110.dll
Up and running !!!
nice :)
/ Reply
Quote
Should RTX 3090 systems be able to download and run TF using Ampere GPUs now or are you still testing? I successfully ran the test app. Do you need the output?
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
Should RTX 3090 systems be able to download and run TF using Ampere GPUs now or are you still testing? I successfully ran the test app. Do you need the output?
nope, should work now.
/ Reply
Quote
It won't give me any work. All I get is this...
(GPU info)
13-Mar-2021 16:59:08 [---] CUDA: NVIDIA GPU 0: GeForce RTX 3090 (driver version 461.72, CUDA version 11.2, compute capability 8.6, 4096MB, 3440MB available, 17791 GFLOPS peak)
13-Mar-2021 16:59:08 [---] OpenCL: NVIDIA GPU 0: GeForce RTX 3090 (driver version 461.72, device version OpenCL 1.2 CUDA, 24576MB, 3440MB available, 17791 GFLOPS peak)
13-Mar-2021 17:02:28 [SRBase] update requested by user
13-Mar-2021 17:02:29 [SRBase] Sending scheduler request: Requested by user.
13-Mar-2021 17:02:29 [SRBase] Not requesting tasks: don't need (CPU: ; NVIDIA GPU: )
13-Mar-2021 17:02:32 [SRBase] Scheduler request completed
Nothing other than WUProp is running on this host.
https://srbase.my-firewall.org/sr5/show_host_detail.php?hostid=209342
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
It won't give me any work. All I get is this...
(GPU info)
13-Mar-2021 16:59:08 [---] CUDA: NVIDIA GPU 0: GeForce RTX 3090 (driver version 461.72, CUDA version 11.2, compute capability 8.6, 4096MB, 3440MB available, 17791 GFLOPS peak)
13-Mar-2021 16:59:08 [---] OpenCL: NVIDIA GPU 0: GeForce RTX 3090 (driver version 461.72, device version OpenCL 1.2 CUDA, 24576MB, 3440MB available, 17791 GFLOPS peak)
13-Mar-2021 17:02:28 [SRBase] update requested by user
13-Mar-2021 17:02:29 [SRBase] Sending scheduler request: Requested by user.
13-Mar-2021 17:02:29 [SRBase] Not requesting tasks: don't need (CPU: ; NVIDIA GPU: )
13-Mar-2021 17:02:32 [SRBase] Scheduler request completed
Nothing other than WUProp is running on this host.
https://srbase.my-firewall.org/sr5/show_host_detail.php?hostid=209342
Do you have enabled AMD and NVidia in pref?
/ Reply
Quote
Do you have enabled AMD and NVidia in pref?
Yes
mfaktc v0.21 (64bit built)
Compiletime options
THREADS_PER_BLOCK 256
SIEVE_SIZE_LIMIT 32kiB
SIEVE_SIZE 193154bits
SIEVE_SPLIT 250
MORE_CLASSES enabled
Runtime options
SievePrimes 25000
SievePrimesAdjust 1
SievePrimesMin 5000
SievePrimesMax 100000
NumStreams 3
CPUStreams 3
GridSize 3
GPU Sieving enabled
GPUSievePrimes 82486
WARNING: Read GPUSieveSize=2047 from mfaktc.ini, using max value (128)
GPUSieveSize 128Mi bits
GPUSieveProcessSize 32Ki bits
Checkpoints enabled
CheckpointDelay 120s
WorkFileAddDelay disabled
Stages enabled
StopAfterFactor class
PrintMode full
V5UserID (none)
ComputerID (none)
AllowSleep no
TimeStampInResults no
CUDA version info
binary compiled for CUDA 11.20
CUDA runtime version 11.20
CUDA driver version 11.20
CUDA device info
name GeForce RTX 3090
compute capability 8.6
max threads per block 1024
max shared memory per MP 102400 byte
number of multiprocessors 82
clock rate (CUDA cores) 1695MHz
memory clock rate: 9751MHz
memory bus width: 384 bit
Automatic parameters
threads per grid 671744
GPUSievePrimes (adjusted) 82486
GPUsieve minimum exponent 1055144
running a simple selftest...
Selftest statistics
number of tests 107
successfull tests 107
selftest PASSED!
ERROR: get_next_assignment(): no valid assignment found in "worktodo.txt"
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
Is an app_config file running?
Use always GPU while computer is in use?
You can also try to reset the project if nothing other WU is running.
Sorry for the questions. Need to track down why its not getting work. If the work cache is full then you will get this message.
/ Reply
Quote
Is an app_config file running?
Use always GPU while computer is in use?
You can also try to reset the project if nothing other WU is running.
Sorry for the questions. Need to track down why its not getting work. If the work cache is full then you will get this message.
It did have an app_config.xml. I deleted it, restarted the client and now I'm getting work. Thanks for the help!
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
Is an app_config file running?
Use always GPU while computer is in use?
You can also try to reset the project if nothing other WU is running.
Sorry for the questions. Need to track down why its not getting work. If the work cache is full then you will get this message.
It did have an app_config.xml. I deleted it, restarted the client and now I'm getting work. Thanks for the help!
great!
/ Reply
Quote
Windows Version + plan_class are done. Lets see how its working with wrapper from mfakto
Update:
The first wrapper doesnt match with the one in download folder, gave some download errors, fixed in v16
Update2:
Doesnt work, same error, need more testing with wrapper.
Update3:
The v12 cuda app is using an older wrapper, will setup this with v15
Update4:
added min_nvidia_compcap to plan_class_spec because of older cards with latest drivers and toolkit (doesnt work with cuda11.2 app)
Question about the Update4 above in bold...
Is that to mean even cards that have newer drivers and Cuda 11 support can not run the v0.18/cuda112 app?
Trying to figure out if its only for RTX3000 series cards or if last gen cards like 2080ti or 1660ti can run on v0.18/cuda112 or not. If they could is there a performance advantage on cuda112 or not? Thanks!
/ Reply
Quote
rebirtherVolunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,691,661
RAC: 42,281
Windows Version + plan_class are done. Lets see how its working with wrapper from mfakto
Update:
The first wrapper doesnt match with the one in download folder, gave some download errors, fixed in v16
Update2:
Doesnt work, same error, need more testing with wrapper.
Update3:
The v12 cuda app is using an older wrapper, will setup this with v15
Update4:
added min_nvidia_compcap to plan_class_spec because of older cards with latest drivers and toolkit (doesnt work with cuda11.2 app)
Question about the Update4 above in bold...
Is that to mean even cards that have newer drivers and Cuda 11 support can not run the v0.18/cuda112 app?
Trying to figure out if its only for RTX3000 series cards or if last gen cards like 2080ti or 1660ti can run on v0.18/cuda112 or not. If they could is there a performance advantage on cuda112 or not? Thanks!
Its a bit complicated, too much drivers but see table here
https://docs.nvidia.com/deploy/cuda-compatibility/index.html#support-hardware__table-hardware-support
/ Reply
Quote