[Martinsliquidlab] Rad & Pump Optimizer V5.6 Spreadsheet - New Parallel GPU feature

[Martinm210]

FYI,
Got a few updates done and now including a parallel GPU feature for you multi GPU users.

Download it here (http://martinsliquidlab.org/pump-and-radiator-optimizer-spreadsheet/).

Quick Screenshot
http://martinsliquidlab.files.wordpress.com/2011/02/martinsradpumpoptimizer56.png

Let me know if anything throws a bad curve. I made quite a few changes and likely made some errors some place.

Cheers!:toast2:
Martin

[zalbard]

Great update, thanks! :up:
Seems to be working as expected. :)

[Church]

Martin: may i suggest to create one thread (or convert this with appropriate changes to it's name) related to spreadsheets like this with subsequent new versions or at least links to their posts/threads or download links to them in OP, so that one can bookmark or subscribe just one thread?

[lowfat]

You have outdone yourself once again.

[miahallen]

Thanks MLL...looks like according to your sheet, the EK GTX 480 FC block is significantly more restrictive than the Apogee XT Rev2. Would you venture a guess as to the difference in restriction between the GTX480 and the GTX 580 (new version) FC block from EK. More or less restrictive?

[Martinm210]

Great update, thanks! :up:
Seems to be working as expected. :)

Thanks!


Martin: may i suggest to create one thread (or convert this with appropriate changes to it's name) related to spreadsheets like this with subsequent new versions or at least links to their posts/threads or download links to them in OP, so that one can bookmark or subscribe just one thread?

Any updates I will make a news post in my blog and update this thread. So you have the option of subscribing at either place.:up:


You have outdone yourself once again.

Thanks! Just trying to keep the science alive...:D


Thanks MLL...looks like according to your sheet, the EK GTX 480 FC block is significantly more restrictive than the Apogee XT Rev2. Would you venture a guess as to the difference in restriction between the GTX480 and the GTX 580 (new version) FC block from EK. More or less restrictive?

I have no idea, but you might ask EK. If you can round up some pressure drop curves from somewhere, I'd be happy to add them in as an [INTERNET] curve.:up:

Substitutions for similar products is always a good place to start. It's not perfect, but neither is the estimation, so it's all good.:up:

[miahallen]

Thanks Martin :)

[CedricFP]

Martin - excellent stuff, as usual. I play with your spreadsheet daily as I fantasise about dream WC setups.

[tiborrr]

Fantastic job, Martin! :up:

[st0ned]

I got a bit strange result with my setup, but maybe it's true ( if it is I really need to revise my pump setup ). I've a ApogeeXT with alt inlet ( because I use 19/13 compression fittings ) and a TPA 120.3 and a TFC 360 with a Laing 655 in the max setting ( = D5 with max setting right ? ). And it only gives me 0.86GPM :( .

Or should I think about buying a new CPU block ? Or maybe a dremel mod for my swiftech top ?

[quattro_]

thanks for all your hard work Martin, much appreciated .

[gmat]

Very neat, i love the parallel GPU option.
I have two D5 with the EK dual top which is not on the spreadsheet, is it the same as if i pick up two separate D5 with EK top V2 ? I suppose it adds them in series ? (pump 1 and pump 2)

[ximius]

Woot! Martin has outdone himself again. Thanks!

[luke997]

Thanks MLL...looks like according to your sheet, the EK GTX 480 FC block is significantly more restrictive than the Apogee XT Rev2. Would you venture a guess as to the difference in restriction between the GTX480 and the GTX 580 (new version) FC block from EK. More or less restrictive?

New EK 580 and 5X0 blocks are much more restrictive.
I went from 3 480 to 2 5X0 and flow dropped a lot. Details on last 2 pages of my worklog.

Upside is the temps are better - obviously :)

[Martinm210]

Thanks!

If the GPUs are restrictive and you have multiple GPU blocks, parallel can provide some huge pressure drop relief. Two in parallel is 1/2 the pressure drop of one...etc. It's pretty hard to choke down a single pump CPU/Multi GPU setup when the GPUs are put in parallel.

Still sporting a stock HSF on my lonely and single 570GTX, but I can pretend now at least..:)

Anyhow, appreciate the comments, glad it's helpful.
Cheers!
Martin

[miahallen]

New EK 580 and 5X0 blocks are much more restrictive.
I went from 3 480 to 2 5X0 and flow dropped a lot. Details on last 2 pages of my worklog.

Upside is the temps are better - obviously :)

Thanks Luke :)

[theseeker]

Thank you Martin! Outstanding as usual.:up:

[bmaverick]

Nice work! :up:

The spreadsheet is really great to see how flow changes in swapping out pumps and RADs to get a good feel.

Love the way it adds up three pumps. Taking three XSPC top DDC-1 9W pumps uses a very small 23.4W of power vs. a three stock DDC-3* at 47.17W. The curves are nearly the same.

[>!eNg_*]

Nice!!
Thank You Martin

[chorner]

Not only is this tool excellent - it shows some mad Excel skills :P

BTW, I found a bug unless I'm not understanding something here. When you add a 2nd GPU, set in parallel it actually increases the loops flow rate... according to the spreadsheet anyhow. That must be wrong, no? When the GPU's are in parallel, it drops the total restriction of having 2 blocks in serial but only a certain %'age. The water still needs to flow in through both blocks... there's no way having two blocks to push water through no matter how it is set up increases your total flow rate. There's something wrong there I think :)