stuck on cuda31

[davidbam]

One machine is not progressing through the usual cuda31 / cuda60 / cuda70 to the higher paying units. I even started again with an older driver to use the same as the others but - no joy. Tearing my hair out on this (if I had any to tear out)

[Dirk Broer]

Not the same that won't run Gerasim? Nope, that was Megacruncher.
Windows or Linux?

[davidbam]

this is Linux sorry. On moo

Now have 3 machines which get only cuda31 WU while all the others progress

[Dirk Broer]

So Moo! on Linux using CUDA. What nVidia card is it that only gets 1.03 WUs instead of 1.04 WUs?
BTW: 1.05 WUs are OpenCL

[davidbam]

These three machines only get v1.03 (cuda31). Doesn't appear to be related to driver version number. All machines with a single GPU do advance past cuda31

Code: Select all

giga2, 192.168.2.214, Linux Ubuntu
Ubuntu 20.04 LTS [5.4.0-33-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9)]

Intel(R) Xeon(R) CPU E5-1620 v3 @ 3.50GHz [Family 6 Model 63 Stepping 2]
8 threads, 10Mb L3cache, 32Gb RAM, 960Gb disk
  ✔avx   ✔avx2   ✔fma
qty 2, GeForce GTX 1660 Ti

Code: Select all

tr2, 192.168.2.206, Linux Ubuntu
Ubuntu 20.04 LTS [5.4.0-33-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9)]

AMD Ryzen Threadripper 2990WX 32-Core Processor [Family 23 Model 8 Stepping 2]
64 threads, 64Mb L3cache, 64Gb RAM, 960Gb disk
  ✔avx   ✔avx2   ✔fma

qty 2, GeForce GTX 1660 Ti

Code: Select all

tr3, 192.168.2.208, Linux Ubuntu
Ubuntu 20.04 LTS [5.4.0-33-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9)]

AMD Ryzen Threadripper 2990WX 32-Core Processor [Family 23 Model 8 Stepping 2]
64 threads, 64Mb L3cache, 64Gb RAM, 480Gb disk
  ✔avx   ✔avx2   ✔fma

qty 2, GeForce GTX 1660 Ti

These two do advance to v1.05 (opencl_nvidia_101) WU

Code: Select all

unk1, 192.168.2.213, Linux Ubuntu
Ubuntu 20.04 LTS [5.4.0-33-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9)]

Intel(R) Xeon(R) CPU E5-2630 0 @ 2.30GHz [Family 6 Model 45 Stepping 7]
12 threads, 15Mb L3cache, 24Gb RAM, 119Gb disk
  ✔avx   no avx2 no fma

qty 2, GeForce GTX 1660 Ti

Code: Select all

asus5r, 192.168.2.182, Linux Ubuntu
Ubuntu 20.04 LTS [5.4.0-33-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9)]

AMD Ryzen 9 3900X 12-Core Processor [Family 23 Model 113 Stepping 0]
24 threads, 64Mb L3cache, 32Gb RAM, 960Gb disk
  ✔avx   ✔avx2   ✔fma

qty 2, GeForce GTX 1660 Ti

[Dirk Broer]

Did you read this posting at the Moo! forum?

Especially the last lines: You need the following files in the project directory, do not forget to chmod +x executables (bold).

* http://moowrap.net/download/dnetc_wrapp ... -gnu__cuda
* http://moowrap.net/download/dnetc520-linux-amd64-cuda60
* http://moowrap.net/download/dnetc-gpu-1.3.ini
* http://moowrap.net/download/job-lin64-cuda60-520.1.xml

But the app_info.xml ain't bad either....

[davidbam]

all happening today = one instance got as far as cuda70 but is now back at cuda60

[bluestang]

all happening today = one instance got as far as cuda70 but is now back at cuda60

OZ/Huey had an issue as well...
https://xtremesystemsteam.org/index.php ... l#msg83937

[scole of TSBT]

all happening today = one instance got as far as cuda70 but is now back at cuda60

Send me the client_state.xml and client_state_prev.xml from that instance

[davidbam]

TIA

emailed