Message boards :
Number crunching :
Observation of CreditNew Impact (4)
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · Next
Author | Message |
---|---|
Wiggo Send message Joined: 24 Jan 00 Posts: 36390 Credit: 261,360,520 RAC: 489 |
Well it didn't take long for Lionel to mass abort another huge amount of MB work, again I see, just so that he can grab more AP work. :-( Absolutely no "Cheers" here. |
rob smith Send message Joined: 7 Mar 03 Posts: 22456 Credit: 416,307,556 RAC: 380 |
Jason - I've just read your thesis on the problems and solutions of CreditNew. And its brought back memories of control theory and its application to industrial scale plant design... I'll have to brave the spiders in the loft to get at my notes. Nice work, and I hope your efforts pay off in getting the beast tamed, its all too easy to have a fast reacting system that is under damped, or a sluggish system that is way over damped - the real skill is getting the system to have the correct combination of damping and gain to keep things in proper control. (Now try it with a four term controller - I've got the right shaped pills somewhere) Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
Thx Jason, You're doing a good job answering our questions that fast. p.s. The next lines are a deep scan from the brain of a person under influence of alcohol: -- clip -- (Understandable words are assembling too but in a hard to decipher order context wisely using You.) I'll feed that input through my analogue inputs. I'm that old that I have studied and calculated that sort of things in electronics rather than in signal processing. But everything I receive is processed "either as positive or even more positive feedback with some needed damping and with the help of some capacitors or coils along with some needed resistors" to do some linear, integral or derivative adjustments to whatever needed depending on the adjustable bias ... :D Or maybe I'll just Crunch. For the F U N of doing that. I'm smiling like an exploded OpAmp or an electrolytic capacitor, (probably due to the 18 proof Winter Beer), when writing this. -- clip ends --- Me, myself and I are going to sleep - unless the other two are snoring. 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 |
Wiggo Send message Joined: 24 Jan 00 Posts: 36390 Credit: 261,360,520 RAC: 489 |
I see that you're aborting MB work again Lionel just to get more AP's. :-( If Credits are what you're after, then why not move to a project that awards more credit p/h. No Cheers. |
Link Send message Joined: 18 Sep 03 Posts: 834 Credit: 1,807,369 RAC: 0 |
I see that you're aborting MB work again Lionel just to get more AP's. :-( (+) |
Blurf Send message Joined: 2 Sep 06 Posts: 8962 Credit: 12,678,685 RAC: 0 |
Shame on you for doing that, Lionel. |
EdwardPF Send message Joined: 26 Jul 99 Posts: 389 Credit: 236,772,605 RAC: 374 |
I couldn't , wouldn't, don't want to, etc read ALL that has been said about "Observation of CreditNew Impact" in all the threads so if this is redundant please forgive me ... It seems the easy way out of this whole controversy is to have Dr.A et al break S@H and AP into 2 projects and let them live separate lives! They can still live on the same computers etc under the covers ... just give S@H and AP space to live on inviable to each other's existence. S@H crunchers VS S@H crunchers for S@H bragging rights and AP crunchers VS AP crunchers dor AP bragging rights. (It can't be THAT hard ... can it??) Ed F |
juan BFP Send message Joined: 16 Mar 07 Posts: 9786 Credit: 572,710,851 RAC: 3,799 |
Or simply as asked a lot of times in the forums, adjust the credit paid by MB to match the same number paid by AP as in theory promised by Creditscrew and never achieved. |
Blurf Send message Joined: 2 Sep 06 Posts: 8962 Credit: 12,678,685 RAC: 0 |
Or simply as asked a lot of times in the forums, adjust the credit paid by MB to match the same number paid by AP as in theory promised by Creditscrew and never achieved. Wash..rinse..repeat. It's up to Dr. Anderson. |
Wiggo Send message Joined: 24 Jan 00 Posts: 36390 Credit: 261,360,520 RAC: 489 |
When the change to SETI MB V7 came into effect AP's here averaged 790 credits per w/u but that is now down to 710 (I only do AP's on CPU) so a leveling is happening again. Remember when AP's use to average 1200 when CreditNew was 1st introduced? Well certain tactics used by some idiots would likely account for that I guess. Cheers. |
juan BFP Send message Joined: 16 Mar 07 Posts: 9786 Credit: 572,710,851 RAC: 3,799 |
Or simply as asked a lot of times in the forums, adjust the credit paid by MB to match the same number paid by AP as in theory promised by Creditscrew and never achieved. Looks like his laundry machine is broken. That´s what i fear... |
cov_route Send message Joined: 13 Sep 12 Posts: 342 Credit: 10,270,618 RAC: 0 |
My suspicion is that the new optimized MB is not as much better than the stock MB as the optimized AP is better than the stock AP. Once our resident Lunatics get MB locked down tight we might see the kind of differential we have been used to, historically-wise. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13835 Credit: 208,696,464 RAC: 304 |
Time to flog the dead horse again. For those of us that crunch just MB going to V7 resulted in my RAC dropping from around 53,000 to less than 32,000. After a while it recovered, sort of, to be around 32,500. Then it dropped down to around 31,000. From that point it gradually climbed back up again to hit a high of 33,500. That was around the begining of November. Since then it has, apart from a minor blip, been falling steadily. It's now down to less than 30,000 (even lower than after the inital introduction of V7) and it is continuing to fall. Why? Grant Darwin NT |
Bernie Vine Send message Joined: 26 May 99 Posts: 9956 Credit: 103,452,613 RAC: 328 |
Well I have been breaking in a couple of new cards in my 3rd and 5th host and I have been logging my top 5 host totals and RAC on a daily basis since 4th of December On the two host with new GPU, there has been an overall RAC increase of No3 Plus 2130 No5 Plus 1122 Much in line with what I had hoped However the other three machines which I have done nothing to over the same 3 week period. No1 minus 1033 No2 minus 281 No4 minus 234 It should be noted that I did this to make number 2 and number 3 identical machines and number 4 and 5 are also now a pair. I don't actually care about RAC except in trying to see how well my machines perform against each other and for my team, I am definately of the school 1 point per WU. Just thought someone might like the figures. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13835 Credit: 208,696,464 RAC: 304 |
I am definately of the school 1 point per WU. If you think people aborting MB WUs to get AP ones is bad, you've no idea how many hundreds of thousands of WUs were aborted in order to get shorties when it was 1 credit per WU. I'm all for the credit reflecting how much work is done. It would just be nice if that actually happened. Grant Darwin NT |
Iona Send message Joined: 12 Jul 07 Posts: 790 Credit: 22,438,118 RAC: 0 |
As I recall from a response I got in an older thread on this subject, basically, V7 is in essence V6 optimised, or something close to that. Grant, I've also seen similar reductions in RAC, albeit on more humble figures. I'm now running my main PC for whole days at a time, to try and claw back some of the loss, rather than for, say, three or four hours plus a few runs overnight and I'm still 'going backwards'! Admittedly, I no longer have the benefit of GPU 'crunching' (Win XP32 and an ATI GPU, limiting useable drivers and 'AMD App', with a reluctance to revert to an older driver and reduce games performance), but I estimate about a 40%+ reduction. As a result, I've often found myself wondering if it is worth continuing, as I can't keep blaming the washing machine for the increase in power use!!! Don't take life too seriously, as you'll never come out of it alive! |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
Hmmm, I see a Barts GPU with an acceptable driver/AMD_APP in Vista. That Host should work fine. Is that the one not working? I don't see a GPU in XP. |
Iona Send message Joined: 12 Jul 07 Posts: 790 Credit: 22,438,118 RAC: 0 |
Unless you have the 'right' OS, driver and AMD App combination, the ATI GPU won't show itself under BOINC and will be 'ignored'. My main is the E8600 with XP32 which had an HD6870 in it, until a few days ago; the 6870 is now 'on loan' to my partner's Q6600 with Vista 64. Previously, I ran an HD4890 under XP32. I then got the 6870 and was seduced by the speed of the thing in games, with Cat 12.3, as opposed to Cat 11.12 used on the 4890 which allowed me to use the GPU for S@H....then found that 12.3 and AMD App, didn't mix under XP32! Still with me? Yep, I basically need to go to something like Win 7 64 bit, so that I can use the newer driver with the respective 'AMD App' as they call it these days (used to be called SDK). Phew!!! Don't take life too seriously, as you'll never come out of it alive! |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
Yes, I'm very familiar with the XP limitations. Actually, unless the Game calls for 12.3, there is very little difference between 12.1 & 12.3. I just picked up a 6870 a couple weeks ago and tested it in my XP host with 12.1. It worked great with 12.1 in XP. Now it's flying high in my other Host preparing to battle for the Highest ranked AMD card in SETI crown. It's scraping paint with a pair of Tahitis at the moment... In Vista, you can install the latest AMD driver, 13.9, and that card should scream. In fact, it would scream in XP with 12.1 just as loud. There are APs available at the moment, that card will complete one in right at 2000 seconds. All the faster Tasks have cleared, but, here's one running the 6870 in XP with Cat 12.1, Task 3278325071 |
arkayn Send message Joined: 14 May 99 Posts: 4438 Credit: 55,006,323 RAC: 0 |
Yes, I'm very familiar with the XP limitations. Actually, unless the Game calls for 12.3, there is very little difference between 12.1 & 12.3. I just picked up a 6870 a couple weeks ago and tested it in my XP host with 12.1. It worked great with 12.1 in XP. Now it's flying high in my other Host preparing to battle for the Highest ranked AMD card in SETI crown. It's scraping paint with a pair of Tahitis at the moment... Lets not forget that it is the top ranked OSX system on SETI. |
©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.