Message boards :
Graphics cards (GPUs) :
6.3.21
Message board moderation
Previous · 1 · 2 · 3
| Author | Message |
|---|---|
[SETI.USA]Tank_MasterSend message Joined: 8 Jul 07 Posts: 85 Credit: 67,463,387 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
true, I do... 100000 to the other 100. I didn't have this problem with 6.3.14 I just uped the buffer time from 1 day to 5 days, and I got a fair amount of new work, no other settings were changed. |
Krunchin-Keith [USA]Send message Joined: 17 May 07 Posts: 512 Credit: 111,288,061 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
For you linux 64'ers try http://boinc.berkeley.edu/dl/boinc_6.3.21_x86_64-pc-linux-gnu.sh version now posted. All versions of 6.3.21 are now avaialble at http://boinc.berkeley.edu/download_all.php?dev=1 |
GDFSend message Joined: 14 Mar 07 Posts: 1958 Credit: 629,356 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() |
6.3.21 is now the suggested client. gdf |
|
Send message Joined: 21 Dec 07 Posts: 47 Credit: 5,252,135 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]()
|
I have a problem which has crept in on a Linux64 host. It tells me there is no work for my platform only atom cell not avail...yada....yada ...however if I switch it back to 6.3.19 it downloads work. This also occurred on the same host when trying 6.3.14,and had to switch back to 6.3.10 to get work. The other thing that happened on 6.3.14 & now again 6.3.21 is screen flicker in the message screen where the alternating lines keep jumping back and forth. My other host shows none of these problems. |
|
Send message Joined: 10 Nov 08 Posts: 4 Credit: 60,350,898 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
On my Kubuntu 64 comp, 6.3.21 doesn't seem to be able to see my GPU. In the message tab it says "no coprocessors found". I've downloaded the latest Nvidia drivers which didn't help. GPU is a 8800GT. I have just installed 6.3.21 on my windows comp & it runs fine there. Only a 8600GT on that one though :-( |
koschiSend message Joined: 14 Aug 08 Posts: 127 Credit: 913,858,161 RAC: 13 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Is the user that runs BOINC part of the group video? BOINC is not in there by default, so it can't use the video card... This message also appears under Linux if you are using the wrong libcudart.so @topic 6.3.21 works very good for me, no problems so far :) |
|
Send message Joined: 10 Nov 08 Posts: 4 Credit: 60,350,898 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Sorry it took me so long to respond, but I've had other Boinc problems & eventually resorted to a re-installation. Originally, I added boinc to the "video" group. This made no difference. Since re-installing I've found that there isn't a user "boinc" to add to the video group??. Should I create one? In the boinc message tab it says "can't load library libcudart", and "no coprocessors found" libcudart.so is in the BOINC folder where the installation process put it, so it should be able to find it. So does anyone have any other ideas :-) |
nutcaseSend message Joined: 16 Oct 08 Posts: 7 Credit: 5,348,057 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]()
|
found a bug in 6.3.21 running a corei7-920 with 2 260 gtx gpu's and XP64 when installed, it defaulted to 10 days cache. changed it to 1 day cache and it still kept trying to get 10 days of work. This caused the other project to run high priority and starved the gpu's of all cpu time. 1 gpugrid wu because of this has been running for over 12 hours and still not completed. |
|
Send message Joined: 17 Apr 08 Posts: 113 Credit: 1,656,514,857 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Been running GPUGrid under 6.3.21 on an i7 for about a week - I have found best WU turn around comes from setting processor usage to 99% under preferences. That way I'm running 7 other WUs and a GPU WU concurrently - GPUGrid WUs are taking about 6.5hours to complete on my 260-216..... Have not seen the 10 day cache issue. |
EdboardSend message Joined: 24 Sep 08 Posts: 72 Credit: 12,410,275 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Been running GPUGrid under 6.3.21 on an i7 for about a week - I have found best WU turn around comes from setting processor usage to 99% under preferences. That way I'm running 7 other WUs and a GPU WU concurrently - GPUGrid WUs are taking about 6.5hours to complete on my 260-216..... ¿How many GPUGRID WU's have you at the same time (crunching + in cache) with your i7? |
|
Send message Joined: 17 Aug 08 Posts: 2705 Credit: 1,311,122,549 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
@Burdett: you's get the same result if you set BOINC to use at max 87.5% of the CPUs, wouldn't you? I'm wondering if there's any unwanted side effect of setting 99% of time.. but there's a separate thread for that. MrS Scanning for our furry friends since Jan 2002 |
|
Send message Joined: 18 Sep 08 Posts: 65 Credit: 3,037,414 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]()
|
@Burdett: you's get the same result if you set BOINC to use at max 87.5% of the CPUs, wouldn't you? I'm wondering if there's any unwanted side effect of setting 99% of time.. but there's a separate thread for that. this percentage thing is totally silly - everyone knows how many cores his machine has. anyone with an X3-phenom abord to check what happens @50%? unless this is meant to be used to control the percentage of shaders to use.. ;) |
|
Send message Joined: 17 Aug 08 Posts: 2705 Credit: 1,311,122,549 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
everyone knows how many cores his machine has. Yeah, I'd say 100% +/- 5 ;) this percentage thing is totally silly Is it? If I want to tell BOINC to run 7 cpu tasks along the one GPU task, I want it to run on 7/8 of all cpu cores, that's 87.5% anyone with an X3-phenom abord to check what happens @50%? The same as if you set between 26 and 49% on a quad core. MrS Scanning for our furry friends since Jan 2002 |
|
Send message Joined: 17 Apr 08 Posts: 113 Credit: 1,656,514,857 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
There are 8 WUs - one running and 7 queued at any one time - which is a little over 2 days work give or take. Sure, 99% gives me the same WU count active as 87.5% .... but I am trying to avoid the 100% (8 other WUs and 1 GPU WU) effect where GPUGrid times go out to 10 hours - and 99% is the first point that achieves that for me. Generally HT has had a net positive effect on the processor to return other WUs, and I am able to hit the optimal turnaround times for my GTX260 at GPUGrid. P. |
|
Send message Joined: 18 Aug 08 Posts: 121 Credit: 59,836,411 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Hi! I am using this version 6.3.21 and it works extremally well. So my question is simple: Is it worth to change from 6.3.21 to 6.5.0 ? Will I crunch faster after change? Now I have drivers: 178.24. Are new drivers better: 180.60 Beta or 180.48 WHQL What you recomend? |
|
Send message Joined: 17 Aug 08 Posts: 2705 Credit: 1,311,122,549 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Is it worth to change from 6.3.21 to 6.5.0 ? No. Will I crunch faster after change? No. Are new drivers better: No. BOINC 6.4.x is recommended more for debug reasons than anything else. There are still problems in these releases and they're difficult to find -> larger user base. The newer drivers are recommended, but as far as I know they're not faster in GPU-Grid. Only if you have the "out of memory" errors with a 64 Bit Win you should definitely upgrade to 180.84. MrS Scanning for our furry friends since Jan 2002 |
Paul D. BuckSend message Joined: 9 Jun 08 Posts: 1050 Credit: 37,321,185 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Hi! Make changes only when needed. If things work well, then there is little or no reason to "upgrade". In my case 6.4.5 was pretty flakey. I wanted to try the GPU processing so went to 6.5.0 ... but, had I not been interested in GPU processing I would have stayed with 5.x.y as it was working well for me. Processing speed is almost completely dependent on the science application. That is the program that does the work. BOINC adds some overhead but it is minimal. |
©2025 Universitat Pompeu Fabra