Originally Posted by ALT F4 Warning: Spoiler! (Click to show)
Originally Posted by Gabrielzm
You can have that with dual d5
Just changed the GPUs in chessboard and have to redo the loop:
GPUs blocks are now in parallel and this new generation of ek blocks (980/titan x) are a bit less restrictive that the old titans blocks. So the end results is if I put both d5 on that loop at 100% I end up having 2 GPM
Before with blocks in serial I was getting 1.4 GPM.
No need to iwaki mate. It will sound like a apartment basement with motors and machines on...@ALT F4
do you have a flow meter on it mate? Do you known how much flow you are getting? One of the best reasons to have a flow meter is to constantly check the health of your loop. If wiuth time the flowdrops you known something is clogging up somewhere... Also how are you calculating the delta air-water? Is the ambient temp measured how? Thermistor close to the case or a separate thermometer in the room but away from the PC?
Your rig is looking awesome
I've answered most of those questions throughout my last several posts. Long story short this room is a solid ~21C to ~23C depending if the air conditioner is on or off. Water temperature will idle at ~29C to ~31C depending on room temp, under load I'm seeing water temp on gpu upwards to ~39C. The room is still cool and you can feel the heat of the furnace errr radiators
I should have videod with the reservoir half full so all could see what I've been mentioning. It's nowhere near a bad flow which is why you would end up confused on why the numbers don't add up
Regardless, when the amd cards are released and I decide what to do with the rest of my rig I might double up on the pumps and call it a day. Next time I test I'm going to use the cards stock bios to see if the 0.05v will make much of a difference, bummer I didn't have enough fittings to test parallel on the gpu loop.
I looked at what pics of your build I could find, and didn't see anything that jumped right out that looked suspect . . .
The acrylic T on the pump outlet is pretty restrictive, I always try to use a triple rotary 90 out of a pump when I need a 90, and then use the straight thru passages of T's for the main flow direction.
I'm not a fan of the type of 90's you have in some locations, particularly the in and out of the GPU bridge.
One thing that you can't see externally, but need to watch for as you build, particularly if you're using fittings from multiple sources, is that some are absolute disasters as far as flow capability . . .
In the pic below, the 45 on the left is a Bitspower, the 45 on the right is some other POS brand, and the M to F rotary in the middle is also some POS brand.
Look at the bore diameters . . . compared to the Bitspower, the others might as well be pinholes.
Just one POS fitting like those two, can trash the flow for the whole loop.
The other thing to watch for as you build, is that some fittings have a longer male thread than others, and some fittings have a very short female thread, that if you use a long male thread into it, can restrict the flow markedly. . . . FC Terminals on GPU blocks would be an obvious example, but surely isn't the only example.
Tearing everything down to inspect stuff is probably the last thing you want to do, but if you fail to get acceptable temp results, you may have to just bite the bullet.