Message boards :
Number crunching :
Linux CUDA 'Special' App finally available, featuring Low CPU use
Message board moderation
Previous · 1 . . . 52 · 53 · 54 · 55 · 56 · 57 · 58 . . . 83 · Next
Author | Message |
---|---|
JohnDK Send message Joined: 28 May 00 Posts: 1222 Credit: 451,243,443 RAC: 1,127 |
In script ... /usr/bin/perl^M ... not found or something Typing whereis perl finds Perl OK perl: /usr/bin/perl5.22-x86_64-linux-gnu /usr/bin/perl /etc/perl /usr/share/perl /usr/share/man/man1/perl.1.gz But I can't find ^M (Ctrl+M) anywhere in the script. On linux, Perl is already installed (type 'whereis perl' to know where the perl binary is located). "#!/usr/bin/perl" in the script seem to be OK, typing perl cpu2gpu.pl -i works: #CPU,GPU,AP wu Tried a simulation run: perl cpu2gpu.pl -r -u GPU version found : 801 - (class : cuda80) So I'm still stuck :( |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
cannot write /var/lib/boinc-client/client_state_new.xml at cpu2gpu.pl line 441.Try running as SuperUser 'sudo' |
JohnDK Send message Joined: 28 May 00 Posts: 1222 Credit: 451,243,443 RAC: 1,127 |
cannot write /var/lib/boinc-client/client_state_new.xml at cpu2gpu.pl line 441.Try running as SuperUser 'sudo' That worked, thanks. Sudo this sudo that... |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
I just leave a SuperUser window (did I say Window? LOL) open so that everything you do is a sudo sudo bash - makes you SuperUser. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
I've searched the thread and didn't come up with an answer to a question I have about what does "Sigma" represent in the stderr.txt output. It varies quite a bit over large values. Is a Sigma value of 3 good? Or is a Sigma value of 400 good? Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
I've searched the thread and didn't come up with an answer to a question I have about what does "Sigma" represent in the stderr.txt output. It varies quite a bit over large values. Is a Sigma value of 3 good? Or is a Sigma value of 400 good? Sigma refers to the width of a Gaussian search. The bigger the value the wider the search under the Gaussian curve (more work). No bad or good meanings. Petri To overcome Heisenbergs: "You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
Thanks Petri. Just wondered what it represented and if it had anything to do with a low value in my case, being represented a lot more in my Inconclusives list. I figured it probably didn't since my increase in the past week was of my own doing. But was curious anyhow. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
W3Perl Send message Joined: 29 Apr 99 Posts: 251 Credit: 3,696,783,867 RAC: 12,606 |
With one sigma, you get 68% of the signal, two sigma 95%, 3 sigma 99.7% and 5 sigma 99.999% In science, most people require a 3 sigma signal to be relevant. In particle physics, they need 5 sigma to make an announcement. |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
Thanks Petri. Just wondered what it represented and if it had anything to do with a low value in my case, being represented a lot more in my Inconclusives list. I figured it probably didn't since my increase in the past week was of my own doing. But was curious anyhow. The value being output is an old debug print out. I was figuring out what the software does with different ar and WU type. Some WU's do not do any Gaussian searches, some do not do any pulse finds with small FFT lengths. Vlars/Guppi do a lot of pulse find work with FFT len 8, 16, 32,..., 1024 and that is SLOW even with the special app. Some workloads spread nicely over all of the GPU compute units, some do not. There is a lot of places for optimisations still to do. That keeps me in this hobby. I would not do this for money. To overcome Heisenbergs: "You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13746 Credit: 208,696,464 RAC: 304 |
There is a lot of places for optimisations still to do. That keeps me in this hobby. I would not do this for money. :-) Reminds me of when I used to repair VCRs, cameras and the like for a living. When my VCR broke down, I didn't repair it for about 3 months, and only then because my brother kept hassling me about it. When you have to do something, it tends to take a lot of the enjoyment out of it. Grant Darwin NT |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
Anybody know if the Special App was tested against the 4-bit WUs in Beta? The reason I ask is that I've had some odd Invalids on two different machines over the last couple days. They're -9 overflows where the Pulse and Triplet counts differ significantly from the wingmen who successfully validated. I'm pretty sure there's also been an increase in my Inconclusives, with a sampling of those looking like these: Workunit 2669316379 (01mr08ac.6680.72.14.41.0) Task 6005913251 (S=0, A=0, P=29, T=1, G=0, BG=0) v8.22 (opencl_nvidia_SoG) windows_intelx86 Task 6005913252 (S=0, A=0, P=0, T=30, G=0, BG=0) x41p_zi3v, Cuda 8.00 special Workunit 2670652172 (10oc08af.29544.82164.10.37.255) Task 6008700184 (S=0, A=0, P=29, T=1, G=0, BG=0) v8.20 (opencl_ati5_SoG_mac) x86_64-apple-darwin Task 6008700185 (S=0, A=0, P=10, T=20, G=0, BG=0) x41p_zi3t2b, Cuda 8.00 special Workunit 2671281097 (14dc08ab.28268.66735.15.42.1) Task 6010004517 (S=0, A=0, P=14, T=16, G=0, BG=0) v8.22 (opencl_nvidia_SoG) windows_intelx86 Task 6010004518 (S=0, A=0, P=0, T=30, G=0, BG=0) x41p_zi3t2b, Cuda 8.00 special Workunit 2671421404 (14dc08ab.30525.69598.16.43.0) Task 6010297125 (S=0, A=0, P=30, T=0, G=0, BG=0) v8.22 (opencl_nvidia_SoG) windows_intelx86 Task 6010297126 (S=0, A=0, P=0, T=30, G=0, BG=0) x41p_zi3t2b, Cuda 8.00 special This may not be limited to Cuda80, though, as I also got an Invalid for a similar Pulse/Triplet mismatch on my daily driver running Cuda50. Only my 4th Invalid on that machine in over 4 years of crunching. |
rob smith Send message Joined: 7 Mar 03 Posts: 22222 Credit: 416,307,556 RAC: 380 |
Hmmm - I'm seeing a number of invalids across my Linux boxes, all of which are running the special app. Some are guppies, but one set are ex-Arecibo, not sue if they were 2 or 4 bit deliveries.... Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
rob smith Send message Joined: 7 Mar 03 Posts: 22222 Credit: 416,307,556 RAC: 380 |
I've just watched a pile of downloads - it would appear that most, if not all, tasks (including ex-Arecibo) we are getting now are 4-bit, despite many being labeled 2-bit Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
I've been seeing a lot of inconclusives too. Additional new ones beyond the ones I admit were caused by me. I am seeing a single count difference either in Triplets or Pulses against my wingmen. They seem to be associated mostly with the new larger tasks. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13746 Credit: 208,696,464 RAC: 304 |
I've just watched a pile of downloads - it would appear that most, if not all, tasks (including ex-Arecibo) we are getting now are 4-bit, despite many being labeled 2-bit From Eric in Beta when they were first released for testing- There won't be any easy way to tell what you've got apart from the file size. 2-bit WUs are about 360K. 4-bit WUs are about 620K. The "2bit" or "8bit" designation in the file name is about the input data that the splitter operated upon. (And it's wrong anyway. The 2bit input data is actually 4 bits per complex sample.) Grant Darwin NT |
rob smith Send message Joined: 7 Mar 03 Posts: 22222 Credit: 416,307,556 RAC: 380 |
The batch I watched were coming in at ~700k rather than the ~360k of late Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
From Eric: Double size is the new standard for both Arecibo and GBT (and eventually Parkes). Unless UC complains about the extra traffic. I don't think they'll even notice it. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
Yes, same here. Based on the FIFO ordering of tasks, I still am processing that big slug of 700K sized tasks. The recent ones today seem to be the older 357K size. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
Strange, my 3x1080 rig does not have one 350k file on it. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13746 Credit: 208,696,464 RAC: 304 |
Strange, my 3x1080 rig does not have one 350k file on it. Likewise, all 705kB files here as well. EDIT- just picked up a 360kB file; resend of a resend. Grant Darwin NT |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.