• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Dual Pump Setup

TekRok

2[H]4U
Joined
Dec 9, 2005
Messages
2,279
I currently have a D5 variant pump pushing through 3 rads, cpu, 2 GPU's, and curious whether Its enough for 2 more GPU's (290) making a total of 4. I have 2 x 480mm rads (XSPC AX480), 1 x 280mm Black Ice GTS, and plan to add a 240mm Alphacool NexXxoS ST30 Rad. Is one D5 able to push through all of that and retain good temps? Or I absolutely need another D5 set in a serial pattern to raise the pressure of the flow?

If one pump is not enough, will this setup work properly by linking 2 pumps from 2 separate points into one another?
42gc.png
 
Last edited:
If you do the gpu's as a block I don't think it significantly increases resistance. If you do them in serial... that does. The radiators have really low resistance, so I don't think adding one more will hurt things that much.
 
You bring up a good point. The cards are currently in a 2 way bridge set up as parallel so I will be setting them up in a 4 way parallel bridge.
 
if you add a second D5, you can have it anywhere in the loop, it doesn't need to be directly after the first D5

as far as whether or not you should add a second D5 based on all of the hardware you're putting into your loop, I personally would add a second D5 (or DDC, or other variant, as serial pumps don't need to be the same), this would allow you to run your GPU's in serial rather than parallel and still maintain a decent L/min & head pressure. it also gives you pump redundancy, though the chance of a D5 failing is pretty slim. running GPU's in parallel *slightly* raises temperatures, so I'd only do it if you have to. I'm running a PMP-500 which is equal to about 2 D5's, but I'm running my cards in parallel due to aesthetics only :)
 
How does running GPUs in parallel raise overall temperatures? It will increase temps of the first GPU and decrease temps of the second GPU as compared to serial, but that's about it. Average temps will remain the same, or be lower due to higher overall flow.
 
How does running GPUs in parallel raise overall temperatures? It will increase temps of the first GPU and decrease temps of the second GPU as compared to serial, but that's about it. Average temps will remain the same, or be lower due to higher overall flow.

if you're running dual pumps, you'll see a marginal decrease in temperatures by running a serial setup, because it's likely you've got enough pressure with two pumps to alleviate the restriction of a serial setup

if you're running one D5 along with a lot of other restriction (multiple radiators, restrictive fittings, chipset blocks, CPU block) most would advise running parallel to maintain flow. MCP-35x or PMP-500 would be a different story, as those pumps are capable of pushing through a lot of restriction before needing a second pump

temps will be very close regardless of which setup you choose

i've always been of the opinion that parallel vs serial is simply based on whether or not you're already hampered by a lot of restriction, or if you're simply fond of how one way 'looks' vs the other

i'm a fan of parallel based on how it looks, so i always use it. temps aren't ever that different either way to really "test" to see which is best. Swiftech did some tests on parallel vs serial and the differences were negligable

if your flowrate is less than 1.5L/min, it could easily raise your temperatures, that's been my experience on my own loops

if your flowrate is 3.5L/min or higher, it's unlikely that going parallel to serial or vice versa would show any difference at all
 
Well, I was talking about a situation where everything was equal. In a situation where everything was equal, serial vs parallel doesn't differ much.
 
From all the research I have done before setting up my system it seems neither parallel nor serial are "better". Parallel provides less flow restriction, and serial provides better cooling in theory since it passes through each card one by one, but it also restricts flow.
 
Back
Top