Message boards :
Number crunching :
Cuda is as Cuda does,,,,,,,,,,,,,,,,,
Message board moderation
Author | Message |
---|---|
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
Sorry......you all know it's f'd...... you all smell like elderberries for trying.........sniff sniff........ "Time is simply the mechanism that keeps everything from happening all at once." |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Just another "Cafe SETI" thread in "Number crunching" forum?.. well-well %) |
Bukken Send message Joined: 3 Apr 99 Posts: 50 Credit: 3,007,776 RAC: 0 |
Well CUDA works for me :-) Might have a few bugs, need to reboot once in a while.. Overall it works.. So thx to Raistmer and everyone else helping fix the CUDA thingy |
RB Send message Joined: 7 Mar 00 Posts: 103 Credit: 1,084,436 RAC: 0 |
..... you all smell like elderberries for trying..... Is that a good thing? |
SATAN Send message Joined: 27 Aug 06 Posts: 835 Credit: 2,129,006 RAC: 0 |
Elderberries are nice, especially elderberry wine. Mark, you need new cat nip, or have you been drinking to much again? |
fjpod Send message Joined: 3 Nov 07 Posts: 57 Credit: 4,919,604 RAC: 0 |
I'm running a Q6700 with an Nvidia 8400GS. I use the Alex Kahn optimized client. Today I Converted it to CUDA. BOINC recognized it and I presume is using it, although I can't tell by looking Processes in Task Manager. I also can't tell by the time per workunit. Seems to be running at about the same speed. What's up with this? |
Iona Send message Joined: 12 Jul 07 Posts: 790 Credit: 22,438,118 RAC: 0 |
I guess this is not a discussion about the Plymouth 440 Cuda, or the like..... Don't take life too seriously, as you'll never come out of it alive! |
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
I'm running a Q6700 with an Nvidia 8400GS. I use the Alex Kahn optimized client. Today I Converted it to CUDA. BOINC recognized it and I presume is using it, although I can't tell by looking Processes in Task Manager. I also can't tell by the time per workunit. Seems to be running at about the same speed. What's up with this? Nobody knows what's up with this............exactly the problem.........if this had been sorted in beta the way is should have been.........no problem here. But I know why it was not so............meow. You will all have to ask the devs......good luck and good night..........see the flick..........'bout the same thing going on here. "Time is simply the mechanism that keeps everything from happening all at once." |
Geek@Play Send message Joined: 31 Jul 01 Posts: 2467 Credit: 86,146,931 RAC: 0 |
The work re-issue rate must be extraordinary. I see a lot of crunchers with hundreds of aborted, detached, and errored out work that must be re-issued. I agree with you Mark. CUDA may be working in the near future but in it's present form it should never had been released. Simply too many errors. Boinc....Boinc....Boinc....Boinc.... |
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
The work re-issue rate must be extraordinary. I see a lot of crunchers with hundreds of aborted, detached, and errored out work that must be re-issued. My main point.......have it work or get it the f outta here.......... I have gotten a lotta crap personally.......... The app sucks....... The way it was brought outta beta sucks....... The lack of addressing the concurrent issues sucks...... And the lack of truth surrrounding why sucks........... Why not come clean, folks??????? Is it that hard to admit a mistake?? Can you hear me now???????????? "Time is simply the mechanism that keeps everything from happening all at once." |
SATAN Send message Joined: 27 Aug 06 Posts: 835 Credit: 2,129,006 RAC: 0 |
When i do crunch with CUDA it's fine. The errors are me ****ing around. It is faster, though with the optimized AP apps things are clearly moving forward. Things will work in about 18 months. Although all the points you have made mark have been perfectly valid. |
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
When i do crunch with CUDA it's fine. The errors are me ****ing around. It is faster, though with the optimized AP apps things are clearly moving forward. Things will work in about 18 months. Duh............ I raise a stink............ The points I make are indeed valid.......... But I still get whacked by those who think they know it all.........you should see me PNs.......... I am about one tid of bowing out of here........and you know that takes a lot......... "Time is simply the mechanism that keeps everything from happening all at once." |
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
(smiles broadly)........you had a 440 cuda??? I had problems keeping tyres on my Buick GS 455....... "Time is simply the mechanism that keeps everything from happening all at once." |
SATAN Send message Joined: 27 Aug 06 Posts: 835 Credit: 2,129,006 RAC: 0 |
You better not go, the doggies will have no one to chase. SEnd me some, I want to see some of them, just to see how stupid they are. |
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
You better not go, the doggies will have no one to chase. SEnd me some, I want to see some of them, just to see how stupid they are.The kitties are about one click outta here......... You stick around and tell me what's going on.......... I have a Nehi rig on the chase that I shall not shut down........ "Time is simply the mechanism that keeps everything from happening all at once." |
.clair. Send message Joined: 4 Nov 04 Posts: 1300 Credit: 55,390,408 RAC: 69 |
That will be the one at `top hosts #21` that I spotted going sonic boom past me, was at #178 Tried the combined MB AP optimised app and things went poop, back on AK v8 MB only now. Its a oc Q6600 for those that want to know. |
Iona Send message Joined: 12 Jul 07 Posts: 790 Credit: 22,438,118 RAC: 0 |
I can quite understand how Mark must feel. As a former engineer, I'm far from happy with what may have passed as QC for CUDA - lots of things just don't make sense. Mark may be a friend and a fellow 'Kitty Person', but that does not mean I have to agree with him! Maybe it is me being 'British' which makes me 'stick up' for him, maybe not. I see a guy who is passionate about what he is doing and wants to see it done properly. Only people who have no feelings towards something, can be dispassionate and unmoved. Mark, quire clearly believes in what he is doing (never mind the outlay of time and equipment) and expresses himself very passionately. I would, too. For Heaven's sake, cut the guy a little slack (as US Marines used to say at RAF St Mawgan - and still do, elsewhere). Don't take life too seriously, as you'll never come out of it alive! |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
I'm running a Q6700 with an Nvidia 8400GS. I use the Alex Kahn optimized client. Today I Converted it to CUDA. BOINC recognized it and I presume is using it, although I can't tell by looking Processes in Task Manager. I also can't tell by the time per workunit. Seems to be running at about the same speed. What's up with this? Mark, Are you seriously saying that it's the project's fault because someone is using Alex's optimized application with an APP_INFO.XML file? SETI@Home is responsible for the override feature through APP_INFO.XML that prevents BOINC from downloading and using the CUDA-enabled science app?? I haven't asked "nobody" specifically, but it seems IOTTMCO to me. As long as he's told BOINC he wants to run AK and not to update, BOINC will not even try CUDA on this machine. Sorry for the rant. -- Ned |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
When i do crunch with CUDA it's fine. The errors are me ****ing around. It is faster, though with the optimized AP apps things are clearly moving forward. Things will work in about 18 months. Mark, I'm not bashing CUDA precisely because I don't have all the facts. From what I understand from Matt's post CUDA is about three percent of all work returned. If we assume that CUDA never validates against the CPU app, and that there is a 0.03*0.03 chance (0.0009, or 0.09%) that two CUDAs will pair up, then the odds of an invalid CUDA result hitting the database is less than 1 in 10,000. The third result has a 97% chance of going to the CPU app. so two CPU apps are very likely to pair up and validate. ... but when CPU and CUDA do not "validate" what that really means is that the results are different. It does not mean that they're not valid. In another post, Matt says they've found the problem in the validator, and it sounds like they're going to be able to go back to the "invalid" CUDAs and revalidate them. In other words, we may be a little bit premature in bashing CUDA. Others have pointed out work in beta that did validate. Someone else could track them down and report if they wish. It is a well-known saying that "Nobody notices when things go right" -- we have to remember that the people who are running CUDA successfully have no reason to complain in the forums. We really can't draw any statistical conclusions from our discussions here because our sample is "self-selected." We're probably better off than reports suggest. -- Ned |
Voyager Send message Joined: 2 Nov 99 Posts: 602 Credit: 3,264,813 RAC: 0 |
Im going to try it when I get a card. Put it on my second machine and see wwhat it does. Actually I've got my cruncher running about as good as can be, so things get boreing. This reminds me of when we switched over from classic, a new boinc version about 2 or 3 times a week.I'm sure the project thinks eventually it will be fine. But I don't like my results getting skewed one way or the other. But time will tell! |
©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.