Posts by alexander

1) (Message 6123)
Posted 1 Jan 2019 by alexander
Is this the new 'me too' ??
2) (Message 6115)
Posted 29 Dec 2018 by alexander
Looks like they are running now. Re-enabled my computers for Asteroid.
3) (Message 6111)
Posted 24 Dec 2018 by alexander
I too have problems with asteroid wu's. They start with an estimated runtime of ~100 minutes, at a runtime of a day or so they remain running @100% . When restarting BOINC they start again at 0%.
For the moment i will delete all wu's, waing for new ones.

4) (Message 5697)
Posted 19 Feb 2018 by alexander
well, yes, i too cannot attach my new tablet (android)
messages says ... get_project_config.php
then it tries to connect to reference site, then it says project may be down.

Sorry, no new connection possible.
5) (Message 5147)
Posted 4 Mar 2017 by alexander
In the messages:
Upload Server out of disk space.
6) (Message 4914)
Posted 25 May 2016 by alexander
upload is working again ..
7) (Message 4906)
Posted 24 May 2016 by alexander
Hello, sorry this project had only one admin, ...

This is why cloning was developed :))
8) (Message 4899)
Posted 24 May 2016 by alexander
This is the moment where we all expect a statement from the admins ...
9) (Message 4058)
Posted 21 Feb 2015 by alexander
I see ~ 20% download errors in my statistics and also crunched but never validated wu's due to too many errors.

I mean, this problem exists now for a really long time. Personal problems or not, it's time to solve them. The project is on risc to loose many crunchers.
10) (Message 4045)
Posted 19 Feb 2015 by alexander
This URL hosts a slightly less than 2 minute video that shows how the Nuclear Test Ban Treaty Organization has detected 26 explosions caused by asteroids in the period from 2000-2013.

Sorry, this link brings me to an 404 error.
11) (Message 4011)
Posted 12 Feb 2015 by alexander
I was not trying to disparage the AMD or ATI GPUs, but was providing a heads-up because other projects had developers who complained about buggy AMD OpenCL drivers which might be solved by passing the correct OpenCL compiler flags. Nvidia's GPUs from Fermi onwards default to IEEE 754 when running CUDA applications and degrade to gaming precision only if the programmer requests that. (Pre-Fermi Nvidia GPUs are stuck at gaming precision and cannot become IEEE 754 compliant in single precision.)
Personally, I think that OpenCL should have defaulted to IEEE 754 compliance and should require developers to explicitly request gaming accuracy to get faster but less accurate results instead of requiring programmers to explicitly request IEEE 754 accuracy to get it, and feel that this is a fault of OpenCL standard instead of AMD. If I had written the standard, such flags would be required to run on hardware that is incapable of IEEE 754 compliance so developers have no illusions on what they are programming.

Hi Jesse,
very interesting info, indeed! Reading statements like your's is one reason why I'm crunching different project, it's much much better than reading newspapers.

I would like to point to a discussion @ GPUGRID, about the new nVidia cards.

12) (Message 4009)
Posted 10 Feb 2015 by alexander
I am going to release new application when I have anything to test on.

Raspberry Pi 2 is available, another source is
13) (Message 4004)
Posted 9 Feb 2015 by alexander
OK, might all be right, but:
Milkyway, GPUGRID, Einstein, Poem, Seti, Collatz and some more do have fine working AMD gpu apps. And in addition, as you posted, these problems do not occur when running double precision. And AFAIK Asteroids is a double precision project.
AMD is well known that they are faster with DP than nVidia. If you do not believe, look into the top hosts list @ Milkyway, a DP project.
Just to give you an impression: the R9 280x is rated SP 4178 GFLOPS, DP 1044 GFLOPS.
A GTX970 is rated SP 3494 GFLOPS and DP 146 GFLOPS !!!
If the project does not need that crunching power - fine, more gpu's left for other projects. Speaking for myself, I'm happy with the cpu apps!
14) (Message 3993)
Posted 8 Feb 2015 by alexander
We have now two volunteers who try to make OpenCL app for AMD cards. Next I'm so busy at work for now. If both of them failed to do OpenCL app, I will do one.


Poem has released a new OpenCL app and has work available 7/24. GPUGRID has a beta app for AMD cards. If you wait too long all gpu's may have found their projects and noone will be here to crunch that ...
15) (Message 3992)
Posted 8 Feb 2015 by alexander
Einstein as well has released an Android5 app.
16) (Message 3967)
Posted 22 Jan 2015 by alexander
Its an FMA4 setup, and I'm watching it sort out its run time right now.

Keep us informed, it may be worth it for more people to use it too!

Hi Mikey,

Well its on its 3rd task, times are around 2hrs 5mins per task:-) That's better than avx or sse.
avx times were 3 hours 20+ minutes & sse were AFAIR 2 hours 30 minutes.

Havent seen any problems as yet, all the tasks completed are waiting validation so I'm hopeful they will do so. Anyway non have failed outright so far.

[edit] 1 validated, 2 pending.. looking good!

Regards, Cliff

I'm using crunch3r's app for many weeks now, not a single one failed or did not validate. It's perfect if your cpu is fma4 capable.
If one wants to check my results look for 'Raj'. The logged errors are all download errors, not crunching errors.
Raj is currently running linux, testing an optimized app in an other project, which seems to be twice as fast as the windows app.
17) (Message 3918)
Posted 4 Jan 2015 by alexander
Thanks for the info and good luck with your new job!
18) (Message 3766)
Posted 10 Nov 2014 by alexander
db_purge was down for three days and yesterday the whole system. Needs some time to re-stabilize. A lot of finished wu's waited for upload and reporting; system was running at 100% and I'm shure a lot of uploads failed, which increased the workload.

Give it a day or so and retry it then.
19) (Message 3755)
Posted 8 Nov 2014 by alexander
Could this bring up a 'deadline' problem?
I have 5 wu's waiting for upload the second day now.
20) (Message 3749)
Posted 4 Nov 2014 by alexander
FYI: My GTX-660 at factory settings lasts an average of 24 seconds before crashing. I have attempted to throttle the GPU usage through the GUI, but as of this time, I can not get it to reduce utilization below 100%.

I have taken GPU processing offline to keep the processor from melting the solder off the board. :)

If this item bug can be fixed easily, please me know. With the advent of newer and faster graphics cards (from multiple vendors) every few months, it probably is not worth the time to investigate this specific issue. IMHO, unless the GPU routines are written for blanket coverage of GPUs keep it at the bottom of whomever's 'ToDo List'. Seems senseless to put a lot of effort into something that would probably work for the next 9 months then need revision(s).

NOTE: I seem to be processing work units fairly well without the GPU. At the time I write this, the main system I have has processed 83,040 work units in 8 days, without using the GPU and while running parallel with SETI
(which I am disabling until at least 2015).

Thanks for listening,


Hi Bridget,
a little bit strange to many people, your wish to bring gpu utilisation down from 100%, most people try to bring it up to this value ...

But step by step: your nVidia card should be able to crunch the gpu wu's, but your driver is not the latest available. But this is not the problem you have.
I see many download errors, they cannot be fixed from your side. Others did not start before deadline, which usually means a too large work buffer. This can and should be fixed on your side. Other errors are not listed in the visible pages of your account.

The other thing is: a@h is a double precision project and the cheaper nVidia cards are not well suited for this type of math, they are very slow. Only the GTX x70 and and higher have better dp performance. AMD would be better, but there is no OpenCL app available.

So for the moment it might be the best advise for you just to disable gpu processing here, the speed gain would be in the range of 2 or less compared to cpu tasks. It's better to use the energy for cpu tasks.

Next 20