Message boards :
Number crunching :
Linux/NVIDIA/AP questions
Message board moderation
Previous · 1 · 2 · 3
Author | Message |
---|---|
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Big Thanks for You Jason for the effort and good job You are doing. We're all waiting for the new improvements. My suggestion for now is to stick with the 'old' (plus your improvements :)). Sorting out x42 is proving to be a monumental effort :) Yeah there's a svn diff command that would compare yout working copy with the original and make a single file with a bunch of +'s and -'s , very handy. If you have the files separate, not under version control, I would do a fresh svn checkout, paste the changed files over the top, then do the diff. It'll work it out exactly as you're thinking. Keep poking at it. You're fiddling in areas I haven't touched since around Fermi introduction, so there's plenty of legroom. The challenge for me now is to make sure I get x42's polymorphic behaviour right, so that hardware that likes it gets it. "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
If you describe what and why changes you did in OpenCL kernels I could imclude those changes in main tree. SETI apps news We're not gonna fight them. We're gonna transcend them. |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
PM coming soon. I send you the AstroPulse_Kernels_r1952.cl I currently run. It has 3 functions/kernels that have changed and they have comments what and why. There may be other minor changes, but diff helps. I have not got the original source and I have lost my benchmark run times, so You may want to bench them first. 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 |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
Hi Guy,
The zip file is quite big. About 150 megabytes. I'll figure out a way to get You access to it. I'll make another zip that has executables, app_info.xml and libraries. Even after that there is a possibility that You can not run them, due to version incompatibilities. I'm running and have compiled all on 64bit Fedora 14, with an updated compiler and Cuda 5.5. 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 |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
No wories Petri, if it's too much trouble, don't worry about it. I'll wait until it gets implemented and distributed through the Seti servers. All I'm lacking, as far as I'm concerned, is a Linux/Nvidia/MB app. I can live with out it for a while, it's not going to reduce my RAC because these are additional PCs to what I'm running now. OK, better to wait. And from what I have understood there are some real advancements in the 'new'. 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 |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14654 Credit: 200,643,578 RAC: 874 |
... Gary Roberts of Einstein is doing some interesting research into GPU crunching on Linux - primarily with AMD/ATI, but I think there may be some crossover here. What caught my eye was ... other things I've read make me think that I don't really need the SDK but rather the 'Installable Client Driver' (ICD) stuff that ends up in /etc/OpenCL/vendors/. Which makes a heck of a lot of sense for an end user. |
Gundolf Jahn Send message Joined: 19 Sep 00 Posts: 3184 Credit: 446,358 RAC: 0 |
chmod'd and chown'd all files after unzipping. boinc:boinc rwxrwxrwx still gives the permission denied error. Did you also change the directories (including the parent)? BOINC needs write access to directories in order to create new files. Gruß Gundolf |
©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.