1) Message boards : Number Crunching : Can validation be changed for inconclusive work units (Message 7408)
Posted 13 Feb 2018 by Alexander
It looks like there is something odd going on.
I cannot upload results, have a lot of errors and do not get new wu's.

Lets hope that the admins will react quick.
2) Message boards : Number Crunching : CNN with OpenCl and Fpga (Message 7367)
Posted 29 Nov 2017 by Alexander
Maybe you have forgotten to say that this card cost 3,990.- USD
3) Message boards : News : [subset_sum] feeder issues? (Message 6418)
Posted 21 Oct 2016 by Alexander
The server issue is still not fixed.

Is there an ETA on resolution?

I am running a Windows Platform, is this the only one affected?

Take a look into server status : the project seems to be shut down.
4) Message boards : News : [dna] new DNA@Home runs up and going (Message 6400)
Posted 15 Oct 2016 by Alexander
this explains why the subset wu's were stopped.
But why the DNA wu's ? Are they also faulty?
I mean, this is a project, driven by a UNIVERSTY, not by a private, who just loves fun. Is it wrong to assume, that formulas are proven to work correct ?
Could please someone, who can do that in perfect english, write a note to the university, asking, why volunteers are sinmply ignored by the admins and request a statement?
And just in case that these servers will simply be shut down, the discussions can be continued at BOINC in the project down section.


5) Message boards : News : [dna] new DNA@Home runs up and going (Message 6395)
Posted 8 Oct 2016 by Alexander

I will shed some light on this. About two weeks ago I started working on SubsetSum app code to add SSE support for it. During my work I discovered serious bug - bit indexing was done in incorrect way (in function all_ones if you want to check this itself). Because of it app generated incorrect results. I wrote email about it to Travis. Week later I sent another one with reminder and details about one more possible bug (this time in shift_left function). This time I got reply that he was busy and will take a look on this. Soon after this work generator was stopped. I thought that Travis will post some official note about this, but for some reason this did not happen.

Thank you Daniel for the info.
Stll not understandable why the project was shut down without any notice.
Anyway, it reminds me on this article:
6) Message boards : Number Crunching : Out of work! (Message 6389)
Posted 1 Oct 2016 by Alexander
No more work available ...
7) Message boards : Number Crunching : relative high error rate (Message 6221)
Posted 1 May 2016 by Alexander

i notice an unusual amount of failed wu's on different PC's. Al my Android wu's validate (except those tried on Android5)
Some examples:

Is there a reason for this?
8) Message boards : Number Crunching : Android wu's fail (Message 6196)
Posted 16 Apr 2016 by Alexander
found the discussion and responded there.
9) Message boards : News : [subset_sum] android app (Message 6195)
Posted 16 Apr 2016 by Alexander
This page @ native boinc
says, all sources are available via github.
10) Message boards : Number Crunching : Android wu's fail (Message 6193)
Posted 16 Apr 2016 by Alexander

process exited with code 1 (0x1, -255)
error: only position independent executables (PIE) are supported.


on this device:
11) Message boards : News : [wildlife,climate,dna,subset_sum] update! (Message 6027)
Posted 30 Nov 2015 by Alexander
Good luck with your new house!
12) Message boards : Number Crunching : No new work! (Message 6020)
Posted 25 Nov 2015 by Alexander
Looks like there is more going on.
Any Info's?
13) Message boards : News : [subset_sum] testing workunits (Message 5063)
Posted 30 Mar 2015 by Alexander

That's interesting! I'll have to look into it. Might be awhile though because I have a new wildlife@home application to release first... too many projects not enough time!

This http://arrayfire.com/getting-started-with-opencl-on-android/ might be a good starting point.
14) Message boards : News : [subset_sum] testing workunits (Message 5057)
Posted 29 Mar 2015 by Alexander
So, I guess I should ask my question a bit different. Any plans to compile your own Android capable app for Subsetsum so the official client can be used? :)

I think the next thing on the agenda for SSS is an openCL version, as I think we'll be able to use GPUs very nicely with this project. As for an android version, it would take either an interested student of mine to work on it (as I've not done any android development) or someone in the community who would want to work with me to port the code over. The application isn't particularly difficult...

Reading these lines an idea came up. For the newer mali gpu's are open-cl libraries available for Android. This gives this project the possibility to become the very first project with an open-cl android app !
15) Message boards : Number Crunching : Computer shuts down (Message 4992)
Posted 10 Feb 2015 by Alexander
I have upgraded to 8GB of RAM and it´s working better now but my computer shuts down sometimes and i don´t know why.

It only happens with this project.

Any clues?

Mem speed setup on AMD-cpu MB is more critical than on Intel-cpu MB.
Check careful the memory timings or, if available, the mem. profile.
Eventually try to reduce the mem clock.
16) Message boards : Number Crunching : Memory usage (Message 4938)
Posted 4 Jan 2015 by Alexander
As a very course calculation I would say:
1 GB for the operating system plus 1 GB per workunit.
17) Message boards : Number Crunching : Memory usage (Message 4933)
Posted 30 Dec 2014 by Alexander
Hi Charles,
thx for your reply.
On one system (with on chip gpu) I had to reduce the max_concurrent even to one. The usable ram is limited there to 3.4GB.

@ Travis:
Now it's clear that this project is not well fitted for the Androids.
These days I installed a low budget system, MB with onboard Intel J1900 Celeron processor. Is a 4 core 2GHz system. Runs with SSD and 100% cpu load and one Einstein Intel gpu wu @ 27watts out of the wall. If someone is interested in the performance: Check my systems, it's Rocky1. It will work 7/24.
18) Message boards : Number Crunching : Memory usage (Message 4930)
Posted 29 Dec 2014 by Alexander
I've noticed on 2 systems, both 2 cores 4 threads, that one Gibbs sampler has paused with the message 'Waiting for memory', leaving one thread unused.
Windows task-manager shows that these processes take more than 900MB of ram. Sorry, this makes it unpracticable on 4GB ram systems, since windows also wants some memory and 4 tasks of this kind are too much. All other apps find 4GB ram enough and I don't want to upgrade to 8GB just for the Gibbs sampler.

Is there a way to reduce the ram usage a little bit? Otherwise I could try to make an app_info that limits the Gibbs-Sampler to a maximum of 2 silmultanous tasks. But this might require some maintenance when the task name changes.

19) Message boards : Number Crunching : Short runs (Message 4853)
Posted 6 Dec 2014 by Alexander
Up- and download time is nearly the same as crunching time. Would it make sense to bundle these short runs to 20 wu's for example in one download?

20) Message boards : Number Crunching : Error with new longer tasks. (Message 4840)
Posted 25 Nov 2014 by Alexander
This task failed after shutting down my pc and restarted:
type_string: reverse, motif_width: 6
starting from checkpoint
reading from samples checkpoint
ERROR: reading samples, reached end of samples before all samples should have been read.
accumulated sample: [279], prev_comma_pos: [558], current_comma_pos: [18446744073709551615]
error on line [270], file [..\..\GitHub\dna_at_home\gibbs_cpp\checkpoint.cpp]

Next 20