• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!
Nov 12, 2024
2
1
15
Hello Everyone,

I’ve had a question on my mind for a long time but haven’t found a solid answer online. To get some clarity, I’m reaching out to both local and international forums that may have the necessary equipment to test this. If no one ends up testing it, I plan to gather the equipment myself and create a video on it.

My question is simple but important: Which cooling solution performs better in long-term, “non-stop 8-12-24 hour” usage: air cooling or liquid cooling? Most comparison videos only show short-term results, but many of us need reliable, long-term performance for gaming or handling heavy workloads (rendering, programming, etc.).

Note: By “long-term usage,” I’m not referring to AIO failure over time; that’s a different topic.

I believe it’s crucial to identify the ideal cooling solution to protect our equipment during demanding new games (like Elden Ring or the upcoming GTA 6) and other CPU-intensive games or tasks.

Cooling Solutions to Test (Examples):

Air Cooling (minimum dual tower): Deepcool, Thermalright, Arctic, Noctua, etc.

Liquid Cooling (minimum 360mm): Deepcool, MSI, Arctic, CoolerMaster, etc.

Test Conditions:

Two identical high-power computers (e.g., AMD Ryzen 7-9 or Intel i7-i9 processors, high TDP CPUs)

Same room temperature: The room temperature for both systems should be identical.

Test Duration: Run Cinebench or similar testing or gaming software continuously for 8-12-24 hours.

Parameters to Observe: CPU temperature, thermal throttling, fan RPM, cooler stability, etc.


Thanks in advance to everyone who participates or shares suggestions!


About Me: I’m a software engineer based in Turkey. I occasionally build powerful systems, use them for a while, then sell them. Although I typically use a MacBook Pro, I switch to Windows for gaming, database management, and cybersecurity work.
 
  • Like
Reactions: Order 66
Hello Everyone,

I’ve had a question on my mind for a long time but haven’t found a solid answer online. To get some clarity, I’m reaching out to both local and international forums that may have the necessary equipment to test this. If no one ends up testing it, I plan to gather the equipment myself and create a video on it.

My question is simple but important: Which cooling solution performs better in long-term, “non-stop 8-12-24 hour” usage: air cooling or liquid cooling? Most comparison videos only show short-term results, but many of us need reliable, long-term performance for gaming or handling heavy workloads (rendering, programming, etc.).

Note: By “long-term usage,” I’m not referring to AIO failure over time; that’s a different topic.

I believe it’s crucial to identify the ideal cooling solution to protect our equipment during demanding new games (like Elden Ring or the upcoming GTA 6) and other CPU-intensive games or tasks.

Cooling Solutions to Test (Examples):

Air Cooling (minimum dual tower): Deepcool, Thermalright, Arctic, Noctua, etc.

Liquid Cooling (minimum 360mm): Deepcool, MSI, Arctic, CoolerMaster, etc.

Test Conditions:

Two identical high-power computers (e.g., AMD Ryzen 7-9 or Intel i7-i9 processors, high TDP CPUs)

Same room temperature: The room temperature for both systems should be identical.

Test Duration: Run Cinebench or similar testing or gaming software continuously for 8-12-24 hours.

Parameters to Observe: CPU temperature, thermal throttling, fan RPM, cooler stability, etc.


Thanks in advance to everyone who participates or shares suggestions!


About Me: I’m a software engineer based in Turkey. I occasionally build powerful systems, use them for a while, then sell them. Although I typically use a MacBook Pro, I switch to Windows for gaming, database management, and cybersecurity work.
IMO, look at what commercial (Dell, HP) servers use. Air for 99%. Only at the truly extreme end does the problems associated with liquid cooling come in.
But servers also use baffles to direct air where needed. This can be done with proper fan placement and balancing.
One aspect you didn't address, was noise. THAT could be the deciding factor.
 
Which cooling solution performs better in long-term, “non-stop 8-12-24 hour” usage: air cooling or liquid cooling?
At high power levels, e.g. 2kW+, server designers are turning to liquid cooling. Not exactly home gaming computer stuff, but it's a big world full of different solutions.

https://www.datacenterdynamics.com/en/news/schneider-launches-liquid-cooled-rack-iceotope-and-avnet/

schneider_iceotope_avnet_liquid_cooled_rack.original.jpg



https://www.iceotope.com/company/ne...d-cooled-8-gpu-server-enabling-ai-everywhere/

gigabyte-g293-render-front-elevated-view-1920x1080.jpg



https://hypertec.com/immersion-cooling/

immersion_cooling_overview_node.jpg



many of us need reliable, long-term performance for gaming or handling heavy workloads (rendering, programming, etc.).
I run long (up to 36 hour) video renders on a 7950X and an NH-D15. A 420mm AIO would probably extract a few more Watts out of the CPU, but I prefer big air. Whatever floats your boat.

My HP servers run screaming Delta fans. Lots of air and deafening noise at startup.


identify the ideal cooling solution to protect our equipment
What makes you think systems need "protecting"? Modern CPUs boost up to the manufacturer's temperature limit under heavy computing loads, e.g. +90°C, +95°C (AMD) or +100°C (Intel) regardless of how "good" the cooling solution is, unless perhaps you're running Liquid Nitrogen or Helium? Perhaps the die still throttles even then.

You could fit an advanced custom loop with a massive radiator or a mid-range air cooler. With any luck the CPU will survive in both cases.

I don't think you'll find an "ideal" solution, but good luck with your endeavours.
 
You're also have to define "room temperature" since most installations that run continuously are installed in specially cooled rooms.
When I refer to ‘room temperature,’ I mean standard home or office conditions—not a specially cooled server room. My goal is to compare the performance of air and liquid cooling solutions during typical, daily usage, like gaming or other high CPU load tasks for 8-12-24 hours straight.
 
You have to be careful about thinking lower temp number is always better.

As long as the temp stays below the point it thermal throttles and reduces the performance a machine the actual number does not matter. A machine running 5 degrees below the max is the same as a machine running 10 degrees below.

The large issue...at least for review publication..is most modern coolers you get can easily cool most cpu. If their test was to say compare the cinnebench number when running different coolers nobody would bother to even read the review when everything gets the same number.
 
Which cooling solution performs better in long-term, “non-stop 8-12-24 hour” usage: air cooling or liquid cooling?
Liquid cooling does produce slightly more heat energy(in watts) than air coolers, due to longer or higher boosts, which in turn mean running air conditioners for longer or even setting them lower.
Is it a fair comparison anymore when liquid cooling releases more heat into the air and increases energy bills more than air cooling?

... many of us need reliable, long-term performance for gaming or handling heavy workloads (rendering, programming, etc.).

Note: By “long-term usage,” I’m not referring to AIO failure over time; that’s a different topic.

I believe it’s crucial to identify the ideal cooling solution to protect our equipment...
That's not a different topic, the failures come with the territory.
If time is money, air is the winner. But if going liquid anyway, then a backup cooler is a must; the downtime from a failed pump or clogged loop can be significant.