Message boards :
Number crunching :
Observation of CreditNew Impact (2)
Message board moderation
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · 9 . . . 20 · Next
Author | Message |
---|---|
ML1 Send message Joined: 25 Nov 01 Posts: 21129 Credit: 7,508,002 RAC: 20 |
... I very much doubt Intel would bother to do that. Much better Marketing to have us have to expensively buy their thousands of instructions of complexity instead! Unless that is AMD fills the niche, or the ASIC or FPGA people blaze a new way... ;-) instruction that can run a typical WU in 60 sec's using the 1 atomic instruction What we have at present is in effect something that averages down to 'option 2' as a greater proportion of new machines use the new super fast instruction. I advocate an 'option 3' and 'option 1' whereby the number of bit flips needed to get the job done on a 'standard processor' is rewarded. Then if you go quantum, you get rewarded with a higher RAC for being more efficient. If you go GPU, you still get a higher RAC for going brute force fast even though only a fraction of your real world bit flips will count. Go ASIC/FPGA for total efficiency :-) Happy super fast crunch in' Martin |
shizaru Send message Joined: 14 Jun 04 Posts: 1130 Credit: 1,967,904 RAC: 0 |
According to Joe (who I'm assuming is an authority on the issue) there is absolutely nothing optimized about the new stock V7 CPU apps. Why do so many of you keep saying it is? It's the same old stock V6 CPU app with a bit of AVX voodoo. |
Keith White Send message Joined: 29 May 99 Posts: 392 Credit: 13,035,233 RAC: 22 |
According to Joe (who I'm assuming is an authority on the issue) there is absolutely nothing optimized about the new stock V7 CPU apps. Why do so many of you keep saying it is? http://setiathome.berkeley.edu/forum_thread.php?id=71728&postid=1371743#1371743 http://setiathome.berkeley.edu/forum_thread.php?id=71773&postid=1373202#1373202 http://setiathome.berkeley.edu/forum_thread.php?id=71960&postid=1379579#1379579 "Life is just nature's way of keeping meat fresh." - The Doctor |
rcthardcore Send message Joined: 23 Nov 08 Posts: 48 Credit: 1,306,006 RAC: 0 |
The new credit scheme is GHETTO! |
Gene F. Send message Joined: 6 Apr 08 Posts: 26 Credit: 2,607,746 RAC: 0 |
It's been fun for the last five years, but it's time to go. The latest version is kicking my poor old 2006 Mac Pro's butt. Fans running fast whenever SETI is working which never happened before. I'll be replacing the machine soon and I'd like to keep it in operation until then. |
Link Send message Joined: 18 Sep 03 Posts: 834 Credit: 1,807,369 RAC: 0 |
It's been fun for the last five years, but it's time to go. The latest version is kicking my poor old 2006 Mac Pro's butt. Fans running fast whenever SETI is working which never happened before. I'll be replacing the machine soon and I'd like to keep it in operation until then. Maybe time to clean them. I'm running v7 on my laptop from 2005, no difference in CPU temperature at all, just like on my desktop system. |
shizaru Send message Joined: 14 Jun 04 Posts: 1130 Credit: 1,967,904 RAC: 0 |
According to Joe (who I'm assuming is an authority on the issue) there is absolutely nothing optimized about the new stock V7 CPU apps. Why do so many of you keep saying it is? Sorry, but that was a lot of work for nothing. The answer is in this very thread:) What part of this sentence... The v6 and v7 CPU apps are about the same speed when doing v6 tasks, with a small advantage to the v7 version on hosts with AVX capability. ...is tripping you up? |
Keith White Send message Joined: 29 May 99 Posts: 392 Credit: 13,035,233 RAC: 22 |
What part of this sentence... And do you realize that v6 and v7 WUs are now built with the same source or is that tripping you up? So of course V6 WU and V7 WU run at the same speed doing the analysis common between the two, it's running the exact same code. Facts The standard CPU app, both the current V6 and V7 app, are using source that includes the Alex Kan optimizations that were the root of Lunatic's optimized seti_enhanced apps of the past. It also now uses an open source FFT library that is AVX aware. Some previous optimized apps used the Intel optimized FFT package which was problematic due to open source licensing problems. I have no idea how the two library's compare in performance but both the standard and current Lunatic's apps now use the same library. The the same analysis is done on a V6 WU as a V7 WU except the V7 includes an auto-correlation analysis pass. "Life is just nature's way of keeping meat fresh." - The Doctor |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
It's been fun for the last five years, but it's time to go. The latest version is kicking my poor old 2006 Mac Pro's butt. Fans running fast whenever SETI is working which never happened before. I'll be replacing the machine soon and I'd like to keep it in operation until then. Just get a couple video cards, mid-range cards work perfectly fine in my upper 4x pcie slots. CPU crunching is so last decade. If I didn't run out of APs, I would be twice as high. Computer 6796479 Look here for an idea of which cards to get, OSX and Graphics Cards Why should I be penalized due to only APs working on my cards? :-) |
Lionel Send message Joined: 25 Mar 00 Posts: 680 Credit: 563,640,304 RAC: 597 |
Can the "scientists" that manage this project explain why "Flopscounter" (under Task details) appears to be showing a lower flops count for v7 work units compared to enhanced work units ... has there been analysis to determine if v7 work units are correctly counting/reporting (I use that phrase loosely) flops. |
Keith White Send message Joined: 29 May 99 Posts: 392 Credit: 13,035,233 RAC: 22 |
Are you talking about the Average Processing Rate? That's a per fix time measurement used to estimate how long a WU is expected to take. <flops> in the app_info does the same thing except APR is dynamic, adjusting for each "completed" WU. The estimated time is used so you don't get over assigned your queue setting for days, assuming you aren't at the 100 WU limit for CPU/GPU. So same flop count divided by longer elapse time equals lower APR. I believe that number is calculated by dividing the average flop count by the average elapse time of valid WUs. But you aren't going to listen to me. "Life is just nature's way of keeping meat fresh." - The Doctor |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13847 Credit: 208,696,464 RAC: 304 |
Maybe time to clean them. I'm running v7 on my laptop from 2005, no difference in CPU temperature at all, just like on my desktop system. Running the new application with the AVX support bumpped by CPU temperatures up by 3-5°c. Grant Darwin NT |
Lionel Send message Joined: 25 Mar 00 Posts: 680 Credit: 563,640,304 RAC: 597 |
Are you talking about the Average Processing Rate? Keith, no I am not talking about average processing rate. |
Link Send message Joined: 18 Sep 03 Posts: 834 Credit: 1,807,369 RAC: 0 |
Maybe time to clean them. I'm running v7 on my laptop from 2005, no difference in CPU temperature at all, just like on my desktop system. That might be because of you are using AVX now, I doubt he is doing that with his 2006 Mac. 3-5°C is also nothing to care about much unless you already had temperature issues anyway. But his description of the problem is just that his fans are running faster than before, that's not an issue actually, not for the machine at least. |
kittyman Send message Joined: 9 Jul 00 Posts: 51477 Credit: 1,018,363,574 RAC: 1,004 |
I am personally tired of this thread. I know, I don't have to read it anymore. I respectfully request that it be closed. It ain't gonna change, and enough is enough. "Time is simply the mechanism that keeps everything from happening all at once." |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
Best evidence is the rewards given my dual card machine. My faster card is consistently given fewer points than the slower card. Aren't they supposed to be about the same? All tasks for computer 6796479 Faster means fewer points? It's broken. Say it ain't so... |
JohnDK Send message Joined: 28 May 00 Posts: 1222 Credit: 451,243,443 RAC: 1,127 |
I am personally tired of this thread. -1 It's not over, I want proper credits for my/our work... ;) |
bill Send message Joined: 16 Jun 99 Posts: 861 Credit: 29,352,955 RAC: 0 |
I am personally tired of this thread. +100 The chances of a response from the 'scientists' that could change anything is some minute fraction of zero. Really, does somebody think they would purposefully expose themselves to the haranguing and abuse that they'd receive from posting in this forum? |
Keith White Send message Joined: 29 May 99 Posts: 392 Credit: 13,035,233 RAC: 22 |
Are you talking about the Average Processing Rate? Then where do you see this ""Flopscounter" (under Task details)" you are talking about. Because I don't see anything in BOINC manager or any of the data driven web pages. I simply want to know what you are looking at and where. "Life is just nature's way of keeping meat fresh." - The Doctor |
Ulrich Metzner Send message Joined: 3 Jul 02 Posts: 1256 Credit: 13,565,513 RAC: 13 |
Then where do you see this ""Flopscounter" (under Task details)" you are talking about. Because I don't see anything in BOINC manager or any of the data driven web pages. It is in every result, look for example here: http://setiathome.berkeley.edu/result.php?resultid=3085140434 (...) SETI@Home Informational message -9 result_overflow NOTE: The number of results detected equals the storage space allocated. Flopcounter: 12763176000.091269 Spike count: 30 Autocorr count: 0 (...) Aloha, Uli |
©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.