Welcome to MilkyWay@home

Posts by Daedalus

1) Message boards : Number crunching : New series of invalids (Message 68984)
Posted 24 Aug 2019 by Daedalus
Post:
Yep, i just discovered it too. Again a "85" run. The other time it was "84" AND "85" runs.

I will crunch some Einstein until they figure out their stuff and communicate clearly that it has been definitively fixed.
2) Message boards : News : Workunit Credit Issues - Fix Implemented (Message 68382)
Posted 24 Mar 2019 by Daedalus
Post:
Hi
Since the server change i get no credits on BOINCstats fpr milkyway WU. What can i do ?


I just checked your account and it seems your last tasks are credited correctly.
3) Message boards : News : Workunit Credit Issues - Fix Implemented (Message 68368)
Posted 23 Mar 2019 by Daedalus
Post:
Thank you for the annoucement. It is now fixed.
4) Message boards : News : Workunit Credit Issues - Fix Implemented (Message 68326)
Posted 21 Mar 2019 by Daedalus
Post:
The amount of credits is lower. Either this is a reevaluation (people won't be happy) or it is another glitch to be corrected in the future. But in any case, make an official annoucement !
5) Message boards : Number crunching : Task Validated But No Credit Given (Message 68322)
Posted 21 Mar 2019 by Daedalus
Post:
I have done a couple WU's this morning and they were credited less than before. They were also a bit shorter. But still less credit for equivalent crunching time.
6) Message boards : Number crunching : Task Validated But No Credit Given (Message 68284)
Posted 20 Mar 2019 by Daedalus
Post:
173566278 1735377989 20 Mar 2019, 3:09:57 UTC 20 Mar 2019, 5:10:09 UTC Terminé et validé 270.37 48.56 0.00 Milkyway@home Separation v1.46 (opencl_nvidia_101)
x86_64-pc-linux-gnu
173566279 1735670078 20 Mar 2019, 3:09:57 UTC 20 Mar 2019, 6:12:09 UTC Terminé et validé 247.82 35.71 0.00 Milkyway@home Separation v1.46 (opencl_nvidia_101)
x86_64-pc-linux-gnu
173404405 1731172266 18 Mar 2019, 2:03:10 UTC 18 Mar 2019, 11:33:33 UTC Terminé et validé 256.45 40.76 227.16 v0.00
173134370 1735650900 15 Mar 2019, 20:38:31 UTC 18 Mar 2019, 3:58:12 UTC Terminé et validé 246.78 39.82 227.17 v0.00
173129546 1735648477 15 Mar 2019, 20:30:28 UTC 18 Mar 2019, 3:49:59 UTC Terminé et validé 246.97 40.06 227.16 v0.00
173123418 1735644708 15 Mar 2019, 20:24:12 UTC 18 Mar 2019, 3:33:12 UTC Terminé et validé 246.72 34.98 227.16 v0.00
173072851 1735625592 15 Mar 2019, 18:56:23 UTC 18 Mar 2019, 2:16:11 UTC Terminé et validé 247.27 39.05 227.17 v0.00
173024825 1735604799 15 Mar 2019, 17:47:28 UTC 18 Mar 2019, 1:07:56 UTC Terminé et validé 248.59 38.14 0.00 v0.00
172920593 1735558868 15 Mar 2019, 15:13:26 UTC 17 Mar 2019, 22:34:59 UTC Terminé et validé 271.64 49.95 243.61 v0.00
172846384 1735525855 15 Mar 2019, 13:27:16 UTC 17 Mar 2019, 20:46:08 UTC Terminé et validé 274.49 53.04 243.61 v0.00
172811393 1735510621 15 Mar 2019, 12:38:06 UTC 17 Mar 2019, 19:57:56 UTC Terminé et validé 262.34 36.39 0.00 v0.00
172725771 1735427255 15 Mar 2019, 10:38:02 UTC 17 Mar 2019, 17:54:14 UTC Terminé et validé 272.75 50.08 0.00 v0.00

I just noticed this: The problem was already there last week. As early as the 15 march. But then it touched only some of the completed units. Now it is all units. I stopped crunching.
7) Message boards : Number crunching : Task Validated But No Credit Given (Message 68283)
Posted 20 Mar 2019 by Daedalus
Post:
Same here. Validated but zero credit.

Plus a lot of validation errors that last week in general.
8) Message boards : Number crunching : New WU Progress Bar (Message 65938)
Posted 21 Nov 2016 by Daedalus
Post:
Anyway i have much less tasks pending for validation.
9) Message boards : Number crunching : Linux and NVIDIA GPU (Message 65929)
Posted 20 Nov 2016 by Daedalus
Post:
Isn't it simply an issue with modprobe ? It is not installed by default if i recall well. My BOINC did not see my nvidia card at first.

If it is the case just type or copy-paste in a terminal:

sudo apt-get update

sudo apt-get-install nvidia-modprobe


If it says it is not in the repository then maybe you need to add the multiverse one.

sudo apt-add-repository multiverse

sudo apt-get update


And then retry to install modprobe.

If you have to manually install openCL maybe installing synaptic form the software center may be handy.
10) Message boards : Number crunching : Underused 970 ? (Message 65569)
Posted 31 Oct 2016 by Daedalus
Post:
Found this thread. Pinned.

http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3118#56955

Added an app_config.xml and got my 970 to crunch two milkyway WU's. I got a meagre 17.65% of output improvement.
11) Message boards : Number crunching : Underused 970 ? (Message 65124)
Posted 9 Sep 2016 by Daedalus
Post:
I used to crunch with an old GTX 460. That toaster did its job honestly. Crunching for example a MW 1.36 in roughly 90 seconds.

Then i got a nice second hand 970 with five times the number of CUDA cores and an advertised quadruple FLOPS output. And i see the 970 crunches a MW 1.36 in 50 seconds. A 80% increase. Nice but below what i expected.

Should i toy with the app_info.xml to make it crunch several WU's at the time ? I don't know at all how to do that. I could toy a bit with the "frame per seconds" parameter too.

I talk about several WU's at the same time because it worked well when i was GPU crunching for einstein.

But what do i know so i wait for you expertise before doing anything stupid.
12) Message boards : Number crunching : All android WU's crashing since a week. (Message 64800)
Posted 4 Jul 2016 by Daedalus
Post:
Ok, answer is there : http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3962
13) Message boards : Number crunching : All android WU's crashing since a week. (Message 64799)
Posted 4 Jul 2016 by Daedalus
Post:
Hello,

I crunch a little bit on two old smartphones. The milkyway 1.00 WU's crash after a second.

My machines:

http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=696647

http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=689564

What should i do ?
14) Message boards : Number crunching : No openCL on nvidia (Linux xubuntu 15.04) (Message 63868)
Posted 12 Aug 2015 by Daedalus
Post:
All right, i found a solution but cannot edit the title to add a [Solved] tag to it.

What to do is described here: https://github.com/bkloppenborg/simtoi/wiki/Installation-Ubuntu-15.04

Googling was not trivial.I copy paste the instructions just in case the page disappears:

1. Prerequisites

The following software must be installed on your machine to build SIMTOI:

sudo apt-get install build-essential g++ cmake libccfits0 libccfits-dev git
sudo apt-get install libglu1-mesa libglu1-mesa-dev
sudo apt-get install opencl-headers ocl-icd-opencl-dev ocl-icd-libopencl1
sudo apt-get install libqt4-dev liblapack-dev libblas-dev libglm-dev

1.1 NVIDIA GPUs

OpenCL on NVIDIA devices requires not only the NVIDIA drivers and OpenCL ICD, but also the Unified Virtual Memory kernel module to operate correctly. All of this software can be installed using the following commands:

sudo apt-get install nvidia-346 nvidia-opencl-icd-346 nvidia-346-uvm nvidia-modprobe nvidia-opencl-dev
sudo reboot


But if you happen to have the problem with an AMD GPU:

1.2 AMD/ATI GPUs

At present Ubuntu does not distribute both the OpenGL and OpenCL drivers for AMD/ATI GPUs/CPUs. To install the OpenGL drivers you can use the package manager:

sudo apt-get install fglrx fglrx-amdcccle

The OpenCL drivers are packaged as part of the
AMD APP SDK. (http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/) Simply download the appropriate compressed file for your architecture, extract it (using bunzip2 or similar commands) and install the resulting files using a command similar to the following:

sudo ./Install-AMD-APP.sh
15) Message boards : Number crunching : No openCL on nvidia (Linux xubuntu 15.04) (Message 63865)
Posted 11 Aug 2015 by Daedalus
Post:
Hello Ladies and Gentlemen,

I used to crunch a few openCL work units on my precedent version of linux (xubuntu 14.04) but now, it seems i have a problem.

I have a nvidia GTX560 and i installed nvidia modprobe. So my GPU is detected and crunches happily einstein units with CUDA. But no milkiway with OpenCL.

My driver is the 346.59. The nivida opencl 346 icd is installed but not the update. Synaptic pretends it has to uninstall the 346 icd to install the update so i am wary of tinkering in my system without knowing what i should do.

lun 10 aoû 2015 21:24:16 CEST | | Starting BOINC client version 7.4.23 for x86_64-pc-linux-gnu
lun 10 aoû 2015 21:24:16 CEST | | log flags: file_xfer, sched_ops, task
lun 10 aoû 2015 21:24:16 CEST | | Libraries: libcurl/7.38.0 OpenSSL/1.0.1f zlib/1.2.8 libidn/1.28 librtmp/2.3
lun 10 aoû 2015 21:24:16 CEST | | Data directory: /var/lib/boinc-client
lun 10 aoû 2015 21:24:16 CEST | | CUDA: NVIDIA GPU 0: GeForce GTX 560 (driver version 346.59, CUDA version 7.0, compute capability 2.1, 1024MB, 981MB available, 1105 GFLOPS peak)
lun 10 aoû 2015 21:24:16 CEST | | Host name: orthanc
lun 10 aoû 2015 21:24:16 CEST | | Processor: 8 GenuineIntel Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz [Family 6 Model 60 Stepping 3]
lun 10 aoû 2015 21:24:16 CEST | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 fma cx16 xtpr pdcm pcid sse4_1 sse4_2 movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm ida arat pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid xsaveopt
lun 10 aoû 2015 21:24:16 CEST | | OS: Linux: 3.19.0-25-generic
lun 10 aoû 2015 21:24:16 CEST | | Memory: 7.75 GB physical, 8.38 GB virtual
lun 10 aoû 2015 21:24:16 CEST | | Disk: 716.27 GB total, 668.50 GB free
lun 10 aoû 2015 21:24:16 CEST | | Local time is UTC +2 hours




©2023 Astroinformatics Group