Folding Forum

Here is some more data on my dual-GTX 660 Ti system, after fixing the downclocking problem:

• cpu: AMD FX-8150 8-core CPU, 3.4GHz (temporarily manually downclocked)
• gpu0: GTX 660 Ti, 1175MHz GPU/6010MHz Memory
• gpu1: GTX 660 Ti, 1228MHz GPU/6010MHz Memory

The GPU WU stats:

• gpu0: PRCG 7623 (314, 1, 8), TPF 5:23, 37697 PPD, 57C
• gpu1: PRCG 7623 (725, 6, 22), TPF 5:11, 39152 PPD, 53C

Total for one SMP and two GPU WUs is 83310 PPD.

compdewd wrote:Having the column headings scroll with the rest of the page will take some JavaScript which I am not fluent in yet, to be honest . I may just look at someone else's code for that and then "copy" it (while learning at the same time, of course).

Here is something to help this:
http://www.cssreset.com/creating-fixed- ... -with-css/

I can look at your code some time and rewrite it to include something that will work to have a scrolling header of the sheet. I am a web developer, but don't have much time to do side projects. I'd be happy to help out and think that I could help you with layout and maybe even some of the DB stuff. Let me know if you're interested.

compdewd wrote:Having the column headings scroll with the rest of the page will take some JavaScript which I am not fluent in yet, to be honest . I may just look at someone else's code for that and then "copy" it (while learning at the same time, of course).

I'm not sure how that might work (JavaScript) if a lot of users have Java disabled in their browsers due to the recent security issues:
U.S. warns on Java software as security concerns escalate

PS: Did you see the entries I made using the Add to The Database form? There were, I think, 6 of them. Thanks for all the great work!

Javascript and java usage are not related. Having java disabled does not affect javascript but some confusion does exist between the two so there are people who might disable javascript which had nothing to do with recent java security problems.

The Java-removal advice doesn't apply to JavaScript

Java is not JavaScript

A common misconception is that JavaScript is similar or closely related to Java.

bollix47 wrote:Javascript and java usage are not closely related.

They are actually not related at all. JS is an in-browser scripting language and has NOTHING to do with any JVM that might be running on the host. Java programs (in bytecode) run directly on the JVM. The only time you would need Java is if a webpage has any applets, and it seems like Compdewds doesn't.

Somebody once said that "Java and JavaScript are similar just like Car and Carpet are similar". It is rather unfortunate that a more distinct name was not picked (if I am not mistaken, the "real" name of JavaScript is actually ECMAScript, but you never really see that used in any broader sense). Stuff like that is just bound to lead to misunderstandings.

Thanks csvanefalk ... wording in post corrected.

Other than the obvious naming of the two technologies part of the confusion exists because javascript can accomplish some of the same functionality that java does thus further causing people to think they're somehow related.

bollix47 wrote:Javascript and java usage are not related. Having java disabled does not affect javascript but some confusion does exist between the two so there are people who might disable javascript which had nothing to do with recent java security problems.

Thanks very much, guys, that's very good to know.

gero1369 wrote:Here is something to help this:
http://www.cssreset.com/creating-fixed- ... -with-css/

I can look at your code some time and rewrite it to include something that will work to have a scrolling header of the sheet. I am a web developer, but don't have much time to do side projects. I'd be happy to help out and think that I could help you with layout and maybe even some of the DB stuff. Let me know if you're interested.

The only problem with doing it with pure CSS is that that would create a fixed header that is fixed no matter what. Since I have filtering options above that actual table header, I can't have the table header fixed as that would cause it to be fixed possibly half-way down the user's screen at all times. The JavaScript code that I am referring to takes the table header with a default CSS property of "position: absolute" and makes it "position: fixed" when the top of the user's screen gets to a certain pixel margin from the top of the web page. I'm going to need a client-side scripting language (JavaScript) to do what I need to do. Also, I have code that works now, but the problem is getting the columns to agree to a width. It is something that will take a little time, which I am currently giving to other aspects of the database. You may PM/email me and we can talk from there.

art_l_j_PlanetAMD64 wrote:PS: Did you see the entries I made using the Add to The Database form? There were, I think, 6 of them. Thanks for all the great work!

Firstly, you're welcome! Secondly, I just now saw them as I have come back from a few days of not having much time. Could you let me know what day you entered those? I have yet to add a timestamp column to the "secondary database" that tells me itself.

As an update to all,
I am currently writing a program that parses only the GPU records from a WuHistory.db3 file from HFM.NET. Then I have to write another program that averages the parsed records. Those are the aspects of the database that I am working on at the moment as I have many records waiting and they will probably take a while.

compdewd wrote:

art_l_j_PlanetAMD64 wrote:PS: Did you see the entries I made using the Add to The Database form? There were, I think, 6 of them. Thanks for all the great work!

Firstly, you're welcome! Secondly, I just now saw them as I have come back from a few days of not having much time. Could you let me know what day you entered those? I have yet to add a timestamp column to the "secondary database" that tells me itself.

OK, it was Friday January 18, when I entered that data.

I have to say again that the database greatly helped me to choose the best GPUs for my criteria of best PPD/$, as described here:

art_l_j_PlanetAMD64 wrote:

compdewd wrote:I'm glad you enjoy the database, Art!

As a reminder, be sure to become familiar with the database disclaimer/waiver before you base purchases off of the database. Thanks!

Thanks, I've just read the database disclaimer/waiver. I did realize, even before I'd read it, that the entries were dependent on the honesty (among other things) of the person making each entry. GPUs with more different entries from different people in the same range of PPDs, I would tend to trust more than GPUs with a small number of entries (like 1 or 2).

But now for some good news I got this morning. An email newsletter from one of my suppliers had a weekly special on GTX 660 Ti's, for $220 each. That changes the $/PPD calculation enormously. I can get either 4 or 5, for a total of $880 or $1100:

• 4 - GTX 660Ti ($220 each) (37k PPD x 4 = 148k PPD)
• 5 - GTX 660Ti ($220 each) (37k PPD x 5 = 185k PPD)

As long as the database user (in this case, me) understands how to appropriately judge the content presented there, and how it relates to the specifications of the GPU they are planning to buy, it is amazingly accurate. For example, I estimated 37k PPD for the GTX 660 Ti, and the actual data for the ones I bought is shown here:

art_l_j_PlanetAMD64 wrote:After the SMP WU was completed, I changed to smp:8 and restarted, here is the data from that configuration:

• smp:8 - PRCG 7809 (2, 5, 104), 3600MHz, 47C, TPF 12:52, 11553.49 PPD
• gpu0 - PRCG 7623 (311, 0, 126), 1175MHz, 57C, TPF 5:22, 37814.76 PPD
• gpu1 - PRCG 7625 (440, 1, 34), 1228MHz, 55C, TPF 5:12, 39026.77 PPD

Total for this configuration is 88395 PPD.

At the risk of sounding like a broken record (that expression really dates me, huh?), thanks again for all the great work!

art_l_j_PlanetAMD64 wrote:At the risk of sounding like a broken record (that expression really dates me, huh?), thanks again for all the great work!

You are quite welcome! Right now, the best and most practical thing I see that I can use my time and skills for is to help the Folding@Home project with this database

My #3 system, AMD Athlon 64 X2 Dual Core 4200+ 2.2GHz CPU, Gigabyte GV-N66TOC-2GD GTX 660 Ti GPU:

• smp:2 - PRCG 6099 (5, 22, 348), 2200MHz, 53C, TPF 1:15:00, 800.87 PPD
• gpu0 - PRCG 7625 (165, 0, 110), 1162MHz, 56C, TPF 5:27, 37236.55 PPD

Total for this configuration is 38037 PPD.


Added the above GPU data to compdewd's Folding@Home GPU Statistics Database, 2013-01-30.

art_l_j_PlanetAMD64 wrote:Added the above GPU data to compdewd's Folding@Home GPU Statistics Database, 2013-01-30.

Thanks for the date I am very busy and still have not got around to adding that date saving feature... Anyway, thanks for the contribution, I will get to adding it within the next couple days.

The latest results for my GPU folding systems:

My #1 system, AMD Phenom II X4 965 3.4GHz CPU, Gigabyte GV-N66TOC-2GD GTX 660 Ti GPU:

• smp:4 - PRCG 8049 (14, 14, 29), 3400MHz, 45C, TPF 1:56, 9218.41 PPD
• gpu0 - PRCG 7626 (11, 0, 91), 1175MHz, 60C, TPF 5:23, 37697.68 PPD

Total for this configuration is 46916 PPD.


My #2 system, AMD Phenom II X4 955 3.2GHz CPU, dual Zotac ZT-40401-10P GTX 460 GPU:

• smp:4 - PRCG 8028 (2019, 0, 57), 3200MHz, 50C, TPF 2:34, 7309.15 PPD
• gpu0 - PRCG 7625 (749, 3, 20), 675MHz, 60C, TPF 11:01, 18421.11 PPD
• gpu1 - PRCG 7623 (98, 0, 138), 675MHz, 61C, TPF 10:52, 18675.39 PPD

Total for this configuration is 44405 PPD.


My #3 system, AMD Athlon 64 X2 Dual Core 4200+ 2.2GHz CPU, Gigabyte GV-N66TOC-2GD GTX 660 Ti GPU:

• smp:2 - PRCG 7200 (0, 2, 466), 2200MHz, 53C, TPF 21:41, 764.98 PPD
• gpu0 - PRCG 7623 (617, 0, 81), 1162MHz, 55C, TPF 5:25, 37465.70 PPD

Total for this configuration is 38230 PPD.


My #6 system, AMD Bulldozer FX-8150 3.9GHz CPU, dual Gigabyte GV-N66TOC-2GD GTX 660 Ti GPU:

• smp:6 - PRCG 7012 (2, 240, 28), 3900MHz, 56C, TPF 3:53, 12287.50 PPD
• gpu0 - PRCG 7625 (98, 2, 50), 1162MHz, 59C, TPF 5:27, 37236.55 PPD
• gpu1 - PRCG 7626 (662, 0, 20), 1228MHz, 56C, TPF 5:13, 38902.08 PPD

Total for this configuration is 88426 PPD.


Total of the above 4 systems is 217977 PPD.

Added all of the above GPU data to compdewd's Folding@Home GPU Statistics Database, 2013-02-03. (YYYY-MM-DD date format)

Reference Only:

System - 1090T stock clock CPU folding SMP with 2 Galaxy GTX460 GPUs set at .825 volts and 525mhz baseclock (derated). Total power consumption 385 watts average. Total PPD ~31k

gpu0 (GTX460) - P8072 48°C 11664 PPD
gpu1 (GTX460) - P8073 50°C 11660 PPD

No intention of turning it up due to older high hour components and cheapo powersupply.

New GPU folding .
GTX 680 direct cu ii.
project 8072.
PPD: 31815. (on stock)
Win 7 64bit
v7