Posts by DeleteNull
log in
1) Message boards : Number crunching : intel A750 GPU Unable to calculate the item. (Message 10423)
Posted 14 Jan 2025 by DeleteNull
I cannot recommend to uninstall "FPGA Emulation Platform for OpenCL" from Intel (https://www.intel.com/content/www/us/en/docs/opencl-sdk/developer-guide/2020-2/intel-fpga-emulation-platform-for-opencl-getting.html) because that will disable other things installed on this computer.

But it seems that mfakto software is confused if more than one OpenCL device is available.
2) Message boards : Number crunching : intel A750 GPU Unable to calculate the item. (Message 10414)
Posted 13 Jan 2025 by DeleteNull
It is this message in your WU with an error:
OpenCL device info
name Intel(R) FPGA Emulation Device (Intel(R) Corporation)
device (driver) version OpenCL 1.2 (2023.16.12.0.12_195853.xmain-hotfix)
maximum threads per block 67108864
maximum threads per grid 0
number of multiprocessors 12 (768 compute elements)
clock rate 0 MHz

Automatic parameters
threads per grid 0
optimizing kernels for INTEL

ERROR: device only supports 0 threads per grid. A minimum of 256 is required for GPU sieving.
23:09:17 (9632): mfakto.exe exited; CPU time 0.000000
3) Message boards : Number crunching : intel A750 GPU Unable to calculate the item. (Message 10411)
Posted 12 Jan 2025 by DeleteNull
It seems you have installed FPGA Emulation

This is the message from one of your failed WU:
OpenCL device info
name Intel(R) FPGA Emulation Device (Intel(R) Corporation)
device (driver) version OpenCL 1.2 (2023.16.12.0.12_195853.xmain-hotfix)

FPGA Emulation will not work with the software mfakto.
4) Message boards : Number crunching : intel A750 GPU Unable to calculate the item. (Message 10395)
Posted 12 Jan 2025 by DeleteNull
Look at here (success):

https://srbase.my-firewall.org/sr5/result.php?resultid=15211668
https://srbase.my-firewall.org/sr5/result.php?resultid=15211764
5) Message boards : News : new TF app testing started (Message 10385)
Posted 11 Jan 2025 by DeleteNull
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?
6) Message boards : Cafe : Rebirther is currently in hospital. (Message 9937)
Posted 30 Apr 2024 by DeleteNull
At least you are in German hospitals, they seems to be somewhat high quality.


For some of the Germans, not for all.
7) Message boards : Number crunching : Dual GPU. Ideas for better cooling? (Message 9929)
Posted 27 Apr 2024 by DeleteNull
You can use a PCI-E extender, so that the GPU have "more air" between them.
8) Message boards : News : Server crash (Message 9928)
Posted 27 Apr 2024 by DeleteNull
Sadly No.

rebirther is currently in hospital.
9) Message boards : Cafe : Rebirther is currently in hospital. (Message 9913)
Posted 17 Apr 2024 by DeleteNull
Hello @all:

Rebirther is currently in hospital. He can write, e.g. in Discord, but is unable to manage the server.

So the upcoming challenge isn't possible.

Recovery hopefully will be possible, but no time schedule yet.

Regards
DeleteNull
10) Message boards : Number crunching : Nvidia Tesla P100 Problems (Message 9761)
Posted 17 Mar 2024 by DeleteNull
That fixed one of my PCs.

Another still errors out on every task
https://srbase.my-firewall.org/sr5/result.php?resultid=145885980


Can you update your driver on this host?
11) Message boards : News : New TF multiGPU apps deployed (issues fixed) (Message 9667)
Posted 7 Feb 2024 by DeleteNull
Are you sure Boinc is in the right order? I see it give a random order compared to say MSI Afterburner. I had a machine with 6 cards in it, and Boinc's order compared to MSI was something like 3, 5, 1, 2, 0, 4. I assumed it was physical connection order vs. driver loading order. Although Folding@Home is different again. Very annoying when trying to allocate cards or work out which is overheating.


Yes in my case with "only" two GPU BOINC has the right order. The GPU where the boot screen appears is GPU0 and the other (dark) one is GPU1. The same order as the BIOS allocates the screens.
12) Message boards : News : New TF multiGPU apps deployed (issues fixed) (Message 9665)
Posted 7 Feb 2024 by DeleteNull
Just made a test for two AMD GPU in a windows host. 5500XT and 6500XT.

The wrapper works well, device 0 has this message:
2024-02-07 13:34:18 (8960): wrapper (7.24.26018): starting
2024-02-07 13:34:18 (8960): wrapper: running mfakto.exe (-d 0)
2024-02-07 13:34:18 (8960): wrapper: created child process 2012
mfakto 0.15pre7-MGW (64bit build)

device 1 has this message:
2024-02-07 13:34:19 (6484): wrapper (7.24.26018): starting
2024-02-07 13:34:19 (6484): wrapper: running mfakto.exe (-d 1)
2024-02-07 13:34:19 (6484): wrapper: created child process 5128
mfakto 0.15pre7-MGW (64bit build)

What does mfakto do?
device 0:
OpenCL device info
name gfx1034 (Advanced Micro Devices, Inc.)
device (driver) version OpenCL 2.0 AMD-APP (3592.0) (3592.0 (PAL,LC))
gfx1034 => 6500XT

device1:
OpenCL device info
name gfx1012:xnack- (Advanced Micro Devices, Inc.)
device (driver) version OpenCL 2.0 AMD-APP (3592.0) (3592.0 (PAL,LC))
gfx1012 => 5500XT

This is a bit confusing because BOINC tells this:
07.02.2024 13:32:57 | | OpenCL: AMD/ATI GPU 0: Radeon RX 5500 XT (driver version 3592.0 (PAL,LC), device version OpenCL 2.0 AMD-APP (3592.0), 8176MB, 8176MB available, 4891 GFLOPS peak)
07.02.2024 13:32:57 | | OpenCL: AMD/ATI GPU 1: AMD Radeon RX6500 XT (driver version 3592.0 (PAL,LC), device version OpenCL 2.0 AMD-APP (3592.0), 4080MB, 4080MB available, 5499 GFLOPS peak)

So this is the opposite assignement to that was mfakto does.

The calculation in fact is done only with one GPU: the 5500XT (e.g. GPU-Z)

So there seems a bug in mfakto, not in the wrapper.

It is not a driver issue, PrimeGrid uses both GPU in parallel.
13) Message boards : Number crunching : Question/Problem with Intel GPU work units (Message 9662)
Posted 6 Feb 2024 by DeleteNull
The main problem is this (mfactc.c):
GPU_type gpu_types[]={
{GPU_AUTO, 0, "AUTO"},
{GPU_VLIW4, 64, "VLIW4"},
{GPU_VLIW5, 80, "VLIW5"},
{GPU_GCN, 64, "GCN"},
{GPU_GCN2, 64, "GCN2"},
{GPU_GCN3, 64, "GCN3"},
{GPU_GCN4, 64, "GCN4"},
{GPU_GCN5, 64, "GCN5"},
{GPU_GCNF, 64, "GCNF"},
{GPU_RDNA, 64, "RDNA"},
{GPU_APU, 80, "APU"},
{GPU_CPU, 1, "CPU"},
{GPU_NVIDIA, 8, "NVIDIA"},
{GPU_INTEL, 1, "INTEL"},
{GPU_UNKNOWN, 0, "UNKNOWN"}
};

Depending on different values for the integer in {GPU_INTEL, 1, "INTEL"}, you will get different results for the number of successful and "not successfull" self tests. But will never get it to 0 of not successful?
14) Message boards : News : A bad day happened (Message 9159)
Posted 15 Oct 2023 by DeleteNull
Just read this in Discord.
Get well soon, and (hopefully) recover all what is broken/damaged!
15) Message boards : Number crunching : Interest in adding a new subproject? (Message 9042)
Posted 15 Sep 2023 by DeleteNull
Just tested. It is slower (NVidia 4080) than a CPU.
16) Message boards : Number crunching : 2 GPUs, 2 tasks on one card, not utilizing gpu 1, (Message 8638)
Posted 7 Jan 2023 by DeleteNull
As far as I know: BOINC passes e.g. "-device 1" to the device so may be we have to update the code that it understands -device instead of -d.
17) Message boards : Number crunching : 2 GPUs, 2 tasks on one card, not utilizing gpu 1, (Message 8636)
Posted 7 Jan 2023 by DeleteNull
Hello, this is the content of a job file:
../../projects/srbase.my-firewall.org_sr5/job_TF_l64c_00020.xml

<job_desc>
<task>
<application>./mfaktc.exe</application>
<append_cmdline_args/>
</task>
<unzip_input>
<zipfilename>mfaktc-linux64-v6.zip</zipfilename>
</unzip_input>
</job_desc>

if you want a device number you have to add a -d <number> parameter (default is 0)

Usage: ./mfaktc.exe [options]
-h display this help and exit
-d <device number> specify the device number used by this program
-tf <exp> <min> <max> trial factor M<exp> from 2^<min> to 2^<max> and exit
instead of parsing the worktodo file
-st run builtin selftest and exit
-st2 same as -st but extended range for k_min/m_max
-v <number> set verbosity (min = 0, default = 1, more = 2, max/debug = 3)
18) Message boards : Number crunching : Trial Factoring (Message 8594)
Posted 30 Dec 2022 by DeleteNull
Hello,

you have to copy the needed libs into your library folder (for me /lib64).
then create a link:
e.g.:
lrwxrwxrwx 1 root root 21 29. Dez 23:46 libcudart.so.12 -> libcudart.so.12.0.107
lrwxrwxrwx 1 root root 21 29. Dez 23:52 libcudart.so.12.0 -> libcudart.so.12.0.107
-rwxr-xr-x 1 root root 675168 29. Dez 23:45 libcudart.so.12.0.107

after this is done mfactc.exe can execute the self test:

Selftest statistics
number of tests 26192
successfull tests 26192

kernel | success | fail
-------------------+---------+-------
UNKNOWN kernel | 0 | 0
71bit_mul24 | 2586 | 0
75bit_mul32 | 2682 | 0
95bit_mul32 | 2867 | 0
barrett76_mul32 | 1096 | 0
barrett77_mul32 | 1114 | 0
barrett79_mul32 | 1153 | 0
barrett87_mul32 | 1066 | 0
barrett88_mul32 | 1069 | 0
barrett92_mul32 | 1084 | 0
75bit_mul32_gs | 2420 | 0
95bit_mul32_gs | 2597 | 0
barrett76_mul32_gs | 1079 | 0
barrett77_mul32_gs | 1096 | 0
barrett79_mul32_gs | 1130 | 0
barrett87_mul32_gs | 1044 | 0
barrett88_mul32_gs | 1047 | 0
barrett92_mul32_gs | 1062 | 0

selftest PASSED!
19) Message boards : Number crunching : Linux GPU errors (Message 7014)
Posted 23 Nov 2020 by DeleteNull

Have you tried running multiple tasks at the same time?


No, I haven't. Performance is outstanding, so there is no need to do so (for me).
20) Message boards : Number crunching : Linux GPU errors (Message 7002)
Posted 22 Nov 2020 by DeleteNull
Depending on the flavour of Linux you are using you have to choose the right directory.
For Ubuntu (and derivatives) it is /usr/lib and in OpenSuse it is /usr/lib64.
But there are some other Linux Distros to confuse the user ;)


Next 20

Main page · Your account · Message boards


Copyright © 2014-2025 BOINC Confederation / rebirther