Most Recent AMD Corsair Hybrid Build - Route and Pump Load

Just a quick inquiry. Threw together an all AMD build with a bit of a mix matched open loop, with some parts I've had on hand after another build I did some time back. However I'm still using hydrox pump and rad (the 120mm one, the 240mm one is swiftech), but went with a parralel route on this build instead of an in-series one, because I liked the idea of both blocks receiving fresh intake right from the pump's run off, and tied in my 120mm rad right at the output of the gpu block, however my CPU block is only running right up to the 240mm rad and then back to the pump. So far so good on temps tbh. I idle and run minimal loads at around 29-31C and then with more moderate to heavy load, jump upward from 32-42C, with around 47C being the highest loop temp I've seen. My CPU temp maintains under 65C and usually runs around 48-52C with moderate load. Attached a breakdown from my iC dash while under moderate load, running blustacks and browsing chrome.

My question is given that the route runs the blocks in parralel with the GPU block being the more heavily heat dispersed, with an extra 120mm rad inline, will I run into any load problems or premature pump failures to maintain my chip performance and temps running the pump the way it is now long term?

I run the pump curve custom, as hydrox config didn't really cut it for me, and my min point on the curve is 2800rpm, with max at 4800rpm. Pump reads fairly steady at 3200-3600rpm to maintain loop temp @ 30-33ish C.

Any particular red flags popping here or am I safe to keep things flowing as is?

My build is in my profile if needed for any reason. Any input is appreciated.

Cheers.

Good to know. Was just a curiousity I had since I have noticed how much higher this build's pump does have to run than other builds i've done and know of. Also, I have been told before that the idea of parallel runs giving the blocks more capacity for heat transfer is much more nominal than the concept would initially have one believe. For me, I think it has more to do with my industry mentality, as I'm a Commercial Refrigeration/HVAC technician by trade, and I often forget not every variant of 'coolant' or substance used to cool goes through a superheating and subcooling processes, the way refrigerant does, so the heat transfer capacity doesn't so drastically diminished as the loop goes on when theres a series of loads when using Water, Gylocol, Coolant, etc. Just a misconception that I think had automatically imprinted on me when i started thinking about how i want to run this particular loop. Aw, well no harm no foul I suppose. lol

Anyhow, thanks for the peace of mind. Much Obliged. Cheers.

With parallel loop, maybe the most important thing to watch is pressure drop across blocks.

Usually when you buy blocks from the same manufacturer, they have roughly the same pressure drop, meaning the flow will split about equally.

In practice, you really need some heavy mismatch to run into problems ^^'

regarding the high RPM, it's typical of parallel loops to be able to have slightly higher pump flow since it can go through the two biggest restrictions in paralle instead of series.

Again.. in practice the most notable difference i saw was less pump noise, maybe because it doesnt push as hard? but it was noticeable in my case.

Pump speed won't have much impact at all. there's a threshold under which the flow will be too small to properly cool the blocks, but beyond that, the gain between this and full speed may be one or two °C.

For example on my particular build, under 1000 RPM, temperatures creep up a lot under load. but between 1200 and 2400 i save 2°C. There's no real need to have a temp curve for the pump imho.

i'd advise to set it to a fixed speed that is sufficient for cooling and that is silent :)

The biggest "problems" i see in your loop is the coolant temperatures. but that's due to radiator space being a bit small, and LL fans being... less than ideal radiator fans.

Also, the 120mm only cools the GPU exhaust and not the whole flow, so you lose a bit more on efficiency there.

Just to give you another example of in/out temperatures from my build, (two 360 and one 240 radiator with fans at full speed), at best i got the output temperature two degrees lower than the input.

under normal use, wether it's heavy load or idle, the temperature difference is under a degree, so don't bother too much about splitting flows across radiators too. send the whole lot through all rads in series if you can.

You'll certainly know that a radiator efficiency increases with temperature difference between ambient and coolant. so the more your loop water rises, the better the rads will be at dissipating heat, even without assisting with fan curves ;)

That's all i can think of wen looking at your build, but that's tinkering to shave every °C possible. in the end, the loop temp is totally dependent on how much rad space you have for a given heat load, and if the case has good airflow. Pump speeds and loop order are mostly cosmetics and noise level preferences.