Overclock.net banner
1 - 20 of 551 Posts

·
Registered
Joined
·
4,474 Posts
Discussion Starter · #1 ·
This is the ongoing and always being tweaked spreadsheet of times and points generated by various projects. I'll update it now and then to improve accuracy, add projects or tasks and make adjustments.

TO VIEW IT, CLICK HERE:

http://spreadsheets0.google.com/spreadsheet/pub?hl=en_US&hl=en_US&key=0AmE73gddEiprdFBHMml3dmxqTjB4T21Mc3Mya3p1Snc&output=html

TO ADD STUFF, CLICK HERE:

https://spreadsheets.google.com/spreadsheet/ccc?key=0AmE73gddEiprdFBHMml3dmxqTjB4T21Mc3Mya3p1Snc&hl=en_US

You can also POST information for me to add! Please feel free to POST, PM, or Edit the Input sheet to add stuff!
biggrin.gif


Added new "New Entries" page and allow anybody to edit it. Any changes on the GPU or CPU pages will be ignored since I maintain the master. Those changes added by OCN BOINC members to the NEW ENTRIES page will be added by me to the main CPU/GPU pages.
06
If you make a new entry to the spreadsheet on the New Entries page, please include OCN name for validation. I'm not going to be tracking down bogus or false entries.

FEEL FREE TO USE ABREVIATIONS! I Know what the project/task names are and can look them up if any questions. MAINLY NEED, GPU/CPU type/speed and TIMES/POINTS!

12/31/11 Happy New Year... added more...
01/12/12 Many additions, updates.
01/22/12 More stuff added...
01/31/12 Added a lot more CPU stuff and some GPU! Thanks all!
02/08/12 Added a bunch, tweak some others...
02/15/12 More GTX470/HD5870 and 2500K stuff..
02/29/12 More CPU and some GPU updates
03/04/12: Minor updates
03/19/12: Minor updates, tweaks...
03/20/12 Added some [email protected] BitCoin GPU #' for 5850
03/27/12 Added some GPU stuff and other task version updates
05/07/12 More stuff, some 7970 and other updates.
05/10/12 Added WCG Beta GPU tasks...
05/12/12 More good stuff!
05/13/12 GTX580 stuff!
06/11/12 More good GPU/CPU stuff
06/15/12 Update GPU/CPU stuff, GTX 680 added
06/28/12 Update GPU/CPU stuff, HD 7950 added
10/09/12 Lot of GPU updates and new 7970 stuff
10/13/12 Add more GTX 580 and some other stuff
10/14/12 Finally got a 7850 in there!
11/14/13 Added more CPU/GPU stuff...
12/4/12 More!
12/9/12 More good stuff!
1/31/13 More good GPU stuff!
2/27/13 Updated a lot of things both GPU and CPU
6/15/13 Just updates for a while... added 7850 XT and more 7970 stuff
3/26/14 Well, added more things, about dang time! LOL! Working on some other spreadsheet ideas as well.

MY FEELINGS ABOUT BOINC....

I BOINC for fun and fellowship with other BOINC'ers world wide to support worthy causes that can benefit all of us. That is enough for me.

The POINTS thing and "combined" points competition is secondary, and fun if not taken seriously. It is however part of the total glue for participation by folks and that is what it's all about! I love and enjoy the talks we all have, the hardware discussions, the strategy talks, all of it; but that is all secondary to just pure participation with whatever one can afford.

In my view, the person running a Pent-M 1.3GHz laptop is doing as much good as one running quad 6990's with dual XEONs. One never knows who will get that special WU that solves a problem somewhere... And, most important, the PERSON, the real life participant behind the Pent-M laptop is no less important than the QUAD 6990 person and should not ever be treated poorly for lack of "ultra high" point generation. People participate within their means and all are a blessing to the whole for that reason. Every participant is important and their personal equipment isn't a factor to me at all.

The ONLY thing "I" really care about is that others care enough to participate. Period. That caring is the stuff that comes from the heart, that compassion, that spark that looks forward and sees the benefits of folding/crunching. All are special in my eyes... with some extremely few special exceptions...

Grandfather, Mother, Father, Sister, Brother all suffered cancer... that why medical projects high priority with me...
 

·
Registered
Joined
·
4,474 Posts
Discussion Starter · #11 ·
Quote:
Originally Posted by DarkRyder;14339058
i get 6700points from 1 work unit on DNetc
wink.gif
takes me 13mins to run them. who rulz?
tongue.gif
Here is what i get... keeping in mind I use Nvidia cards... Keep in mind the cards are used in pairs... DNETC uses both GPU's at the same time. SLI is disabled in all systems, but in the past, that made no difference in computing other CUDA tasks.

GTX 560 Ti SC 950MHz-
10.27Min, 816 points, Task 1.02 [email protected] (cuda31)

GTX 460 794MHz-
11.18Min, 816 points, Task 1.02 [email protected] (cuda31)

9800 GT 734MHz-
27.23Min, 816 points, Task 1.02 [email protected] (cuda31)

I would say your one 6990 beats all mine added together... Guess I have to find one and put it on water and let'r rip...

biggrin.gif
 

·
Registered
Joined
·
244 Posts
Good luck finding a 6990 tex.
Has any one been using moowrapper? I've been using it in conjunction with dnetc. I believe they are both the same thing afterall.
 

·
Jeepin
Joined
·
242 Posts
Looking over your spreadsheet I see your 9800gt does Prime Grid the same as my 6850, about an hour per wu. I can't believe the 6850 does that bad, is the cuda platform that much better, or is the project just optimized for nvidia chips?
 

·
Registered
Joined
·
244 Posts
Quote:
Originally Posted by 4x4n;14346252
Looking over your spreadsheet I see your 9800gt does Prime Grid the same as my 6850, about an hour per wu. I can't believe the 6850 does that bad, is the cuda platform that much better, or is the project just optimized for nvidia chips?
I think the prime grid is just that optimized for cuda.
frown.gif
 

·
Registered
Joined
·
4,474 Posts
Discussion Starter · #17 ·
I agree, some apps work better with Nvidia, some with ATI. All apps that use double precision math will run faster on ATI parts since GeForce products don't do native DP (64b) math, only single precision math (32b). DNETC is one of those ATI optimized apps and screams through DP tasks...

I don't know specifics about PrimeGrid, but it likes CUDA and runs fast on Nvidia stuff...

biggrin.gif
 

·
Registered
Joined
·
4,474 Posts
Discussion Starter · #19 ·
You know something, the CUDA vs. the ATI Stream processor argument will never end and is a moot point. Saying one is more optimized doesn't really tell the story since doing projects for them requires the Manf SDK. Taking it a step further, one could say the SDK's are more optimized... again a moot point.

Whatever the projects use, however the code is written, we are stuck with it. The results and speeds are what they are and it makes no difference what folks say the reason is. At some point, hardware DOES make a difference at some level... The way Nvidia treats DP math by converting it to Float in the compiled code works... but ATI doesn't need to convert and can run 64bit DP math natively. I won't get any deeper into hardware internals because it is again a moot point.

WU's are what they are and run the way they run... speculatiing on the reasons why one runs fast or not is useless. All we need is the base results to compare since we all run the same tasks.

So, that is why I put up the spreadsheet. Today, I will make it open to editing so others can add their results if they wish, but will always maintain a local MASTER copy just in case. I would ask folks use the THIRD "New Entries" page to input their numbers and I will later move them to the other pages.

biggrin.gif
 
1 - 20 of 551 Posts
Top