thanks for the thorough information, this helps us to understand what's going on.

+1. I didn't know the QC development situation.
Another solution is to open the QC code (if it is possible) to community.
TN-Grid, for example, had a lot of help from two volunteers that helped the team to introduce the sse/avx extensions

TN-Grid, for example, had a lot of help from two volunteers that helped the team to introduce the sse/avx extensions

I think on every project where I have seen sse/avx, etc. it has been the volunteers that do it. It takes special expertise. Also, there has been at least one project (POEM, if you remember it) where a volunteer did an outstanding job with an OpenCl app for GPUs, though I don't have his name at the moment. And a very capable developer has done both CUDA and OpenCl apps for XANSONS for COD.

If you ever visit Rosetta, you will quickly find out that rjs5 knows a whole lot about extensions and parallelism; I have seen him at Einstein too.
If you put out a call, they will come.

If you ever visit Rosetta, you will quickly find out that rjs5 knows a whole lot about extensions and parallelism;

I know Rjs5 (i chatted privately with him time ago).
He have a deep knowledge about c, c++ and optimizations.

If you put out a call, they will come.

I don't know. For example, in Rosetta, the developers seems to be not interested in optimizations. (and, also, they have restrictions on open the code).

Actually the current beta QC version should not be too different from the final one. One common reason for failure is that the computations need huge temporary files. We can't make them need less space, but the disk limit will be raised (this occurred for the linux app too).

I just got another test WU - and again, it failed with the
"196 (0xc4) EXIT_DISK_LIMIT_EXCEEDED" error:
http://gpugrid.net/result.php?resultid=20424952

So, obviously the disk limit was not raised enough :-(

Actually the current beta QC version should not be too different from the final one. One common reason for failure is that the computations need huge temporary files. We can't make them need less space, but the disk limit will be raised (this occurred for the linux app too).

Will these be released again for Linux? What was the limit raised to?

I installed 4TB HDD into 3 machines in anticipation of these tasks. Would like to see how they run.

---

I installed 4TB HDD into 3 machines in anticipation of these tasks. Would like to see how they run.

Me too. I am putting 500 GB on all my new machines (and some old ones too). If that isn't enough, I can do more.

How much disk space do the QC tasks actually need, on the average?

@[VENETO] boboviz: The code is open source https://github.com/psi4/psi4/

I've got 2 linux machines running QC tasks.

My 40 thread machine is running 10 tasks at once with 4 threads per task. According to boinc, GPUGrid is using 18.8 GB of disk space.

My other machine is running 4 tasks simultaneously at 4 threads per task and that one is taking 9.19 GB of disk space.

The linux app (3.31) is running very smoothly these days. Kudos to Toni et al for fixing the bugs of the earlier versions.

@[VENETO] boboviz: The code is open source https://github.com/psi4/psi4/

Great!!!

About a month ago, Stefan wrote:

We have stated this in the past but I'll explain it here again. GPUGRID is related to the lab. In the lab we don't have a C++ developer, they are all chemists, biologists and statisticians. Best bet would be me as a computer scientist but I've been busy with many other things recently. So the development of the Windows app has fallen on Raimondas who is very nice to do this as he works in a company (Acellera) which obviously has clients to take care of and does not have an immediate gain from making this work on GPUGRID and Toni who tests it out of personal interest. The current paper we are working on takes priority so once it is finished which is supposed to be this week Raimondas and Toni might finish up the app. According to their latest news the app works mostly fine now except lacking some optimization so it should not be too far off.
Patience please, we will make it.

My posting now is not to push or pester, just asking for information on the current status of the project.
Many thanks in advance -

I highly doubt anything has changed as they still don't have a C++ developer. I would just lay low and let it happen when it happens.

I highly doubt anything has changed as they still don't have a C++ developer.

There is no student/graduate student/Phd/etc in University able to develop in C++?
I'll wait for windows app, in meantime i crunch with my linux vm

Got this error message on the current batch of QC test for Windows:

3/28/2019 7:47:29 AM | GPUGRID | Aborting task 1445r30-TONI_TST11-0-1-RND5643_0: exceeded disk limit: 44800.51MB > 28610.23MB

And another:

3/28/2019 9:07:38 AM | GPUGRID | Aborting task 1445r7-TONI_TST11-0-1-RND1466_0: exceeded disk limit: 44800.51MB > 28610.23MB

Please let us know when this is fixed and I will start up some more test tasks.

For now it does not seem a widespread problem. Can you reset the project?

Did a project reset and started three more tasks. Two task finished successfully and the other aborted with:

3/28/2019 10:37:48 AM | GPUGRID | Aborting task 170r5-TONI_TST11-0-1-RND6393_2: exceeded disk limit: 34955.30MB > 28610.23MB

As a separate test, I tried the tasks on a different Windows PC number 489422. All four tasks aborted with the following message:

<core_client_version>7.14.2</core_client_version>
<![CDATA[
<message>
(unknown error) - exit code 195 (0xc3)</message>
<stderr_txt>
10:28:17 (15156): wrapper (7.9.26016): starting
10:28:17 (15156): wrapper: running install_miniconda.bat ()
10:28:18 (15156): install_miniconda.bat exited; CPU time 0.000000
10:28:18 (15156): app exit status: 0x1
10:28:18 (15156): called boinc_finish(195)

I did a project reset and got the same result. Am I missing something?

Can you reset the project?

I try on a new machine (Intel Xeon and Windows 7)
All errors:

17:51:36 (6372): install_miniconda.bat exited; CPU time 40.357459
17:51:36 (6372): wrapper: running .\qmml3\python.exe (run.py)
Traceback (most recent call last):
File "C:\ProgramData\BOINC\slots\0\qmml3\lib\site-packages\psi4\__init__.py", line 55, in <module>
from . import core
ImportError: DLL load failed with error code -1073741795

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "run.py", line 14, in <module>
import psi4
File "C:\ProgramData\BOINC\slots\0\qmml3\lib\site-packages\psi4\__init__.py", line 60, in <module>
raise ImportError("{0}".format(err))
ImportError: DLL load failed with error code -1073741795
17:51:37 (6372): .\qmml3\python.exe exited; CPU time 0.000000
17:51:37 (6372): app exit status: 0x1
17:51:37 (6372): called boinc_finish(195)

I try on a new machine (Intel Xeon and Windows 7)
All errors:

That's strange. No errors on my other machine (other Xeon) with Win10 (1809 version) 64 bit.