Would it be beneficial to set up two separate water loops for multiple GPUs to enhance overclock performance?
Would it be beneficial to set up two separate water loops for multiple GPUs to enhance overclock performance?
The Delta T must occur or there will be no cooling. It occurs at the radiator water to air boundary. This emphasizes the need for strong airflow. In one loop, you must bring down all components to CPU temperature thresholds. This demands greater radiator area and increased airflow compared to separate loops for the CPU and GPUs. If only GPUs are looped together they can operate at higher limits with less cooling needed. The CPU can then be cooled within its safe range. The 1*C temperature difference in the coolant loop prevents this in a single loop.
I'll try to explain. when your CPU is at full load and runs let's say 75C, the coolant will be about 40C and the air temp would be 25C.
So the delta T is there and there is nothing wrong with it (component - coolant, coolant - air are way more than 1C). and yet, since in a "healthy" loop the coolant moves fast - really fast, the temperature of the coolant is almost the same across the whole loop. If you slow down the flow, the differences might get larger, but that makes cooling much less efficient. It's (not so) simple thermodynamics.
So if you have a loop with CPU and 2 GPUs with let's say 6x120 rad surface, it will cool all the components same or better than if you split it to 2x120 and and 4x120. And the reason for that is simple - a rad can dissipate a certain amount of heat with some airflow at a given delta T (coolant - air). The cooling capacity of the rad scales linearly with the surface - read double the rad size, you can handle double the heat at the same airflow and delta-T. And the components (CPU and GPU) also has this parameter of how much heat they can transfer to the loop at specific delta T (component - cold plate of the block). That's why it's completely normal to see different components at different temps (for example CPU at 75C and GPU at 45C in the same loop).
There are several factors that influence delta-T in a loop: thermal load, coolant flow rate, airflow over heat exchangers, and the surface area of the exchangers. Also, since this hasn't been discussed before, I would vote against using an aluminum radiator if possible. You're still considering a 240+360 radiator in terms of surface area. Additionally, when you refer to 'best performance,' could you clarify what you mean? Are you talking about cooling efficiency, overclocking potential, or overall component effectiveness? Running a single loop is generally the most effective approach, though it's important to note that the pump model wasn't specified. I would suggest sticking with one loop and running GPUs in parallel rather than in series.
the only sensible reasons to have separate loops are:
1. to enhance liquid flow when the pump is insufficient.
2. for simpler handling of swamping parts.
now regarding your loop, replace the aluminum rads or use aluminum for everything else.
aluminum doesn’t work well with copper or brass; mixing them can cause rapid galvanic corrosion.
after considering this, for a 2x980ti you should have at least 4x120 rad surface (more is ideal) and around 2x120 rad for the CPU loop. The rads perform better when in one loop instead of multiple.
OP, I think this advice is solid. Also, Aurora is a show coolant and likely not the best option for regular use, particularly if this is your first loop.
https://mayhems.co.uk/mayhems-aurora-guide/
Thanks for the help, now I understand why my previous Ryzen 7 build’s coolant all Aurora pearls disappeared probably got stuck in the radiator...
n0ns3ns3 :
I'll try to explain. when your CPU is at full load and runs let's say 75C, the coolant will be about 40C and the air temp would be 25C.
So the delta T is there and there is nothing wrong with it (component - coolant, coolant - air are way more than 1C). and yet, since in a "healthy" loop the coolant moves fast - really fast, the temperature of the coolant is almost the same across the whole loop. If you slow down the flow, the differences might get larger, but that makes cooling much less efficient. It's (not so) simple thermodynamics.
So if you have a loop with CPU and 2 GPUs with let's say 6x120 rad surface, it will cool all the components same or better than if you split it to 2x120 and and 4x120. And the reason for that is simple - a rad can dissipate a certain amount of heat with some airflow at a given delta T (coolant - air). The cooling capacity of the rad scales linearly with the surface - read double the rad size, you can handle double the heat at the same airflow and delta-T. And the components (CPU and GPU) also has this parameter of how much heat they can transfer to the loop at specific delta T (component - cold plate of the block). That's why it's completely normal to see different components at different temps (for example CPU at 75C and GPU at 45C in the same loop).
Thanks Will & Nonsence & Rubix for the details.
Well i just notice that i mistaken some info whereby I'm having 2 x 360mm rad + 1 x 120mm rad, basically I'll put it like back 120mm + top 360mm + front 360mm.
But still I'm not comfortable to put a 360mm rad at the front reason being currently I'm having a Ryzen 7 build which having the front with a 360mm rad as well. I i notice the air it pulls into the case is fairly warm which bugs me since I'm afraid it will cause some bad temp to my components (M.2 SSD especially). Take note my environment temperature is around 29 degree Celsius - 34 degree Celsius since i don't have any aircond at my place.
So now let's get back to the suggestion you guys provided:
- No no for mixing aluminum & bronze - Must follow
- Still considering back 120mm rad + top 360mm rad for 2 GPU & front 360mm rad for CPU how does this sound?
- Loop wise will be 2 individual loop since I'm having 2 pump + 2 temp monitor device as well
Airflow wise every single rad will be push pull setup which means 360mm rads get 6 fans & 120mm rad get 2 fans
Sample of my existing Ryzen 7 loop (All Aurora pearl gone & disappear)
7700K Build my 1st hard tubing build with back 120mm rad + top 360mm rad + side 240mm rad
Don't fret about the air intake through the rad. It will slightly increase the internal temperature of the case, but not significantly. SSDs behave differently—while the controller on an SSD doesn’t overheat easily, the memory chips actually perform better and last longer at higher temperatures. If you're concerned, there are heatsinks available for SSDs.
I also live in a room with around 33°C most of the year, and both front and top rads are pushing air into the case—everything is fine. The ideal setup would have all components drawing air from inside, but that’s not possible here because of limited space.
The push/pull method seems unnecessary and wasteful, especially with thin rads and low fin density. If you're using 30-40mm thick rads with around 15FPI, it's better to use a pull configuration.
The reason for pulling is mainly for easier cleaning. In push or pull setups, dust, hair, or debris tend to build up between the fan and the rad, eventually blocking airflow. You’ll then need to disassemble the fan to clean it, which might require removing the radiator from the case.
About two loops, I’d still recommend a single loop if you’re concerned about noise. A 360mm is too large for a CPU while a 360+120mm is barely sufficient for a 2x980Ti. The math shows that each 120mm section handles around 150W at ~2000RPM with decent fans. Overclocked 980Ti can easily exceed 300W under load. So, a 360+120 setup is probably too loud and inefficient.
For CPU cooling, every 120mm section usually provides about 150W of heat dissipation at around 2000RPM with good fans. A 360mm would be more than enough for a GPU while only slightly affecting the CPU during heavy loads. In gaming or CPU-intensive tasks, the GPU will draw more power, so a single loop is often better.
You might want to check if radiators work well with 140mm fans—around 280mm offers similar performance. Using a 420mm radiator (three times 140mm) would greatly improve cooling. Alternatively, opting for a single loop could be a good choice.
I still believe a single loop remains the optimal choice in this situation. There aren't strong examples showing that two loops outperform a single one for this configuration. I've tested heavily overclocked Q6600 with SLI GPUs using a single loop at 2x360 rads, and it performed well with a single D5 pump. Most people overlook the fact that minimizing restrictions within your loops can allow you to run a single large loop with excellent flow rates and stable temperatures if done properly. Coolant flow rate plays a major role, which is why many opt to split loops instead of creating one optimized loop.
I've decided on the following adjustments:
- Radiators will be positioned at the top and front sides
- Fans will all be pushed in for efficiency
- I'll clean the filters weekly, including the front top and inner areas of the PC
- Two GPUs will use back and top radiums while the CPU uses front radiums due to lower heat output
- I'm aiming to reuse existing components instead of purchasing additional sizes (sticking to budget limits)
- The design will follow a single loop with the simplest possible routing
- The current plan includes Reservoir Pump → GPU2 → GPU1 → 120mm rad → 360mm rad → CPU → 360mm rad → Reservoir Pump
your plan seems sensible.
a few notes:
1. even with regular filter cleaning, dust will eventually accumulate between fans and radiators after some time.
2. regarding loop order, placing the CPU immediately after the GPUs might be beneficial. this helps prevent fluctuations during the loop setup, making filling and draining smoother. imagine opening the drain port or valve—you want the liquid to exit naturally.