credit comparison to other projects
Message boards :
Number crunching :
credit comparison to other projects
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 . . . 15 · Next
| Author | Message |
|---|---|
 Pigu Send message Joined: 22 Mar 08 Posts: 3 Credit: 57,884,314 RAC: 0  
|
50/h but on which cpu? 
|
|
Send message Joined: 21 Aug 08 Posts: 625 Credit: 558,425 RAC: 0 
|
50/h but on which cpu? Yes, therein lies the rub... As I said, I get about 50/hr on an overclocked Athlon64 3700+ (clocked to 2750 MHz with DDR-500 memory, compared to the default of 2200 MHz with DDR-400). This same system gets *about* 25/hr over at Einstein. The Einstein tasks have a runtime variation that causes some fluctuation above and below that value, but it is a fair "average"... As you scale up and down in CPU technology and frequency, you will get different cr/hr (or CPCS, Credit Per CPU Second) values, so just saying "50 per hour" is somewhat meaningless without the context...that's why I said that the most amount that the project should reduce to is half of the current credit award. Phrased that way, it would cut it in half across all CPUs. However, for credit consistency I personally feel that no change should be made. Around 50/hr was what I was getting with the 1.22 application, so if I'm getting around 50/hr now, that means that except for the short period of time that Milksop's app was in use, historical credit values are comparable inside this project... |
Labbie Send message Joined: 29 Aug 07 Posts: 327 Credit: 116,463,193 RAC: 0  
|
I get about 1 credit per minute for both MW and SETI (AK_V8 App) on my Q9450 @ 3.36 GHZ.  Calm Chaos Forum...Join Calm Chaos Now |
|
Send message Joined: 14 Jul 08 Posts: 50 Credit: 8,398,033 RAC: 0 
|
Until you can have an app thats stable across all OS's. Why not wait it out for a week or so and see how many errored out wu's will drop the cr per hr. Then make an adjustment. Your moving way too fast... Slow down and see what the numbers really are across all O.S.'s. |
|
Travis Send message Joined: 30 Aug 07 Posts: 2046 Credit: 26,480 RAC: 0  |
Until you can have an app thats stable across all OS's. Why not wait it out for a week or so and see how many errored out wu's will drop the cr per hr. Then make an adjustment. Yeah, I haven't made any changes yet, but wanted to see how people were doing with the new app. 
|
|
Send message Joined: 21 Nov 08 Posts: 90 Credit: 2,601 RAC: 0  
|
Until you can have an app thats stable across all OS's. Why not wait it out for a week or so and see how many errored out wu's will drop the cr per hr. Then make an adjustment. My Machine's very happy with the 0.6 app. |
|
Send message Joined: 14 Jul 08 Posts: 50 Credit: 8,398,033 RAC: 0
|
Until you can have an app thats stable across all OS's. Why not wait it out for a week or so and see how many errored out wu's will drop the cr per hr. Then make an adjustment. Yea, well good for you, then you have check pointing, I don't, I still have yet to see the 0.6 app. Another point to wait, the new app is not even applied across all O.S.'s yet. |
caspr Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0
|
With the 04 app. I'm making 20CPH on a 3.0 xeon xp64 pro A clear conscience is usually the sign of a bad memory  |
|
Send message Joined: 21 Aug 08 Posts: 625 Credit: 558,425 RAC: 0
|
I get about 1 credit per minute for both MW and SETI (AK_V8 App) on my Q9450 @ 3.36 GHZ. The tasks you're using to compare with at SETI that yield about 1 cr/min (60/hr) are VLARs (Very Low Angle Range) which will have lower payouts. You have at least 3 other clusters of pay ranges there on that same system: 70-75/hr 80-85/hr 88-92/hr The other "full disclosure" piece about this is that over on SETI you're using a fully optimized application. I don't know how it relates to the application here, but to me it looks like your average cr/hr there is about 72-76, which means that if MW is averaging you 60/hr, then from your perspective, MW is already "paying less" than SETI, despite the chart at BOINC Combined Statistics claiming otherwise... Edit: BOINCStats has this for CPCS for that system at SETI: Average credit per CPU second 0.020041 0.020041 * 3600 = 72.1476 My 72-76 was not bad for just eyeballing it... ;-) Here: Average credit per CPU second 0.028518 0.028518 * 3600 = 102.6648 Since it is so close to 108, my guess is that the average is going down from the artificial limit that the project enforced. It needs some more time to decay, then what should happen is the value for SETI will actually be higher... That's what I called the "loophole" for SETI. So long as the other projects compare against that composite stat at BOINC Combined Statistics, you're never really going to have "parity for everyone", only "parity on average" |
Arion Send message Joined: 10 Aug 08 Posts: 218 Credit: 41,846,854 RAC: 0 
|
50/h but on which cpu? On my slowest system Windows XP Pro with an AMD Athon XP 3000 (2.1 GHz) stock I am getting about 42 credits for 58 minutes. On fastest system Vista 32 bit with an AMD Athon X2 6400 (3.2GHz) stock I'm getting about 39 credits per 40 minutes. Looks like on that one with a long and a short its just about 50 credits per hour. I don't remember what it was before Milksap's client, but with his I was limited to 108 an hour across the board. I was also getting about 5 minute wu's with his app. Now that we have changed to the new MW app I'm seeing increased times, apparently more science and about 1/2 the credits. Also I'm not waiting endlessly for new work and banging the servers in the process. Looks like a win/win situation all around as far as I'm concerned. I'm fine with the way things are at this point and have no complaints. 50 credits an hour seems reasonable based on what I've seen. If all platforms are running the new stock app for MW and getting about 50 credits per hour then I don't see any reason for anyone to complain. This is based on that its fairly distributed amoung all platforms. Someone tosses in an optimized client and then we're back where we started. I also have 2 other computers that are inbetween my fastest and slowest shown and ALL are running 50/50 with other projects.  
|
Kevint Send message Joined: 22 Nov 07 Posts: 285 Credit: 1,076,786,368 RAC: 0 
|
50/h but on which cpu? A Pent II of course !!! |
DoctorNow Send message Joined: 28 Aug 07 Posts: 146 Credit: 10,703,601 RAC: 0 
|
50/h but on which cpu? I'm getting nearly 50 credits/hour on my AMD X2 5200 under Windows 32-Bit, too. Would do a bit better under Linux 64-Bit I guess, but I haven't tried it out yet, my Linux-hard disk is a bit corrupt and I haven't repaired it yet. ;-) Member of BOINC@Heidelberg and ATA! My BOINCstats |
|
caspr Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0
|
50/h but on which cpu? Don't just think Linux is faster than Windoz! 64 bit windoz on Milsops app. did a WU in 10 mins. and it took 12.5 mins on Linux 64,.... same box, dual boot, 3.0 Xeon,...'03 server With DDR2-400 Ecc Reg's A clear conscience is usually the sign of a bad memory |
|
[B@H] Ray Send message Joined: 27 Dec 07 Posts: 35 Credit: 1,432,926 RAC: 0 
|
50/h but on which cpu? How many per hour and my old 486 SX 25, or the old XT? I have not started them up in years. 
|
Cori Send message Joined: 27 Aug 07 Posts: 647 Credit: 27,592,547 RAC: 0 
|
50/h but on which cpu? True! Linux 64 bit is slower than Windoze 64 bit! So credits are at ~49cr/h on my Windows lappy but only ~39cr/h on my Linux quad. Lovely greetings, Cori  
|
|
Send message Joined: 16 Nov 07 Posts: 23 Credit: 4,774,710 RAC: 0 
|
My 2 laptops, C2D T5600 and T7250, both run XP 32 bit. On both the new MWAH Optimized rewards roughly 80% of Seti Multibeam Optimized (AK v5.13 SSSE3). It's a bit on the high side, but acceptable in my opinion. MWAH credits on both these lappys are about 1.7 times the benchmark values. On my ancient (7 years now) P4 1.8 XP 32 bit it's about 2.7 times. Although benchmarks tend to vary a lot (BOINC CC version, OS, the weather outside, mood of the electricity used etcetera) in my experience the benchmark values are about right for these three comps. The P4 getting rewarded much more could have something to do with the math being used at MWAH. AFAIK the P4 is much more suited for integer computations than for floating point, eg. P4's are very useful for running PrimeGrid. Just my EUR 0.03. |
 Send message Joined: 21 Jul 08 Posts: 2 Credit: 25,054,554 RAC: 0
|
Think I must be doing something wrong then, 41 credits an hour on Seti with the optimised app, 33 credits an hour with the new app here. Intel dual core @ 2.1GHz. |
|
DoctorNow Send message Joined: 28 Aug 07 Posts: 146 Credit: 10,703,601 RAC: 0
|
Think I must be doing something wrong then, 41 credits an hour on Seti with the optimised app, 33 credits an hour with the new app here. Intel dual core @ 2.1GHz. Wow, what optimized app do you use for SETI? I never got a credit level better than 30 credits on my AMD with the app I used... Member of BOINC@Heidelberg and ATA! My BOINCstats |
speedimic Send message Joined: 22 Feb 08 Posts: 260 Credit: 57,387,048 RAC: 0
|
Think I must be doing something wrong then, 41 credits an hour on Seti with the optimised app, 33 credits an hour with the new app here. Intel dual core @ 2.1GHz. There you go: ...on my AMD... mic. 
|
|
DoctorNow Send message Joined: 28 Aug 07 Posts: 146 Credit: 10,703,601 RAC: 0
|
There you go: ...on my AMD... Is there no better optimised SETI-app for an AMD? That's pretty unfair in my eyes... ;-) Member of BOINC@Heidelberg and ATA! My BOINCstats |
©2024 Astroinformatics Group