Asetek Reports Lower Q3 2025 Revenue Due to Fewer Liquid Cooling Products Shipments

[-]

InflammableAccount@reddit

AIOs = not performant enough to warrant the high prices, unless you have a 13/14th gen Intel K. Doesn't last forever.

Custom water: Can last many times more than AIOs, modular parts, best performance in almost any application, costs as much as a used car or new motorcycle.

Air: $50, almost as good as any AIO, lasts forever.

[-]

Zen5 with AVX512 is already so demanding, it's frying the exactly great and not well cooled VRMs. Cooling the CPU itself is still not that bad due to the lower thermal density of relatively distant chiplets, but it's already pushing the comfort limits of non-industrial users of air cooling.

Zen6 rumors make me believe that AMD is going to release furnaces. If you just go with the idea of Zen5 tightly packed together, with more cores and less bottlenecked by an outdated I/O die, that's already hot.

Air cooling is still going to be okay for the gaming crowd not even buying the top CPU models, and never moving their setups, making the massive hanging tower less of a problem.

Also, is it just me, or custom water loops don't actually last that long if you mean time to required maintenance, not time to critical failure? I really got tired with regular maintenance because of course the loop already has a lot of questionable goods in it after a year, and the additive manufacturer points at the tube likely leaching plasticizer into the loop, the tube manufacturer doesn't even point back just demands its own premix liquid to be used, and silly pointing games go on forever.

I occasionally wonder why don't we have hybrid coolers working as standalone air coolers with an optional liquid chamber for additional cooling, but then likely it's because it couldn't be marketed as better than AIOs, and most people don't even touch the hardware after a specific setup is assembled.

[-]

Llew19@reddit

Lol I had a custom loop for a 5820k and two 970s and after three years suddenly thought 'oh no, I've not done any of the maintenance people go on about' .......I changed the fluid. That was it, even with clear tubing. TBF this was back when EK wasn't the shit show it is now, but I'm not sure what all the fuss is about. PC was fine for another 3 or 4 years before being replaced entirely

[-]

ComplexEntertainer13@reddit

I've not done any of the maintenance people go on about

Really boils down to what you do with the loop.

If you are running some fancy colored liquid or are one of the "distilled water is fine" crowd, prepare to do maintenance.

Meanwhile the recommendation is to replace your DP-Ultra ever 2 years or something. But that recommendation is based around worst case loop setups when you actually deplete the corrosion inhibitors in a mixed metal setup for example. I have had that shit in rigs for 3-4 years and blocks looked like new when retiring them.

[-]

Green_Struggle_1815@reddit

"distilled water with no addatives is fine"

i ran that for years without issues.

[-]

AntLive9218@reddit

This is the uncertainty that rarely go addressed.

I've also had such a setup for a while, and it had no problems.

A different setup with a car coolant additive that's often not recommended was also okay.

Another setup with a recommended additive ended up with odd blobs in the reservoir.

Yet another setup with different parts but yet again with a recommended additive ended up with black spots growing inside one specific kind of tubing with other parts and other tubes seemingly unaffected.

It's great when there's nothing going wrong, but there's no guaranteed recipe for getting that even for all the money thrown at the problem, so I started to understand the AIO crowd more after all the encountered issues.

[-]

Green_Struggle_1815@reddit

I agree. The issue is we don't know how reactive the setup will be because we can't really analyze the materials electric potentials etc. Thus having a better inhibitor always helps. Seeing that even AIO manufactures put additives in, even though they can control the variables to a far higher degree than we can, it's likely the sensible move to add them in a custom loop as well.

[-]

Green_Struggle_1815@reddit

the cooling solution nearly doesn't matter. the cpus spike to 100C near instantly given a truly heavy load.

[-]

InflammableAccount@reddit

Dunno what warm VRMs have to do with AIOs vs. Air. vs Custom-water.

If anything, there are only a few coolers on the market that make a notable difference in VRM temps: Air coolers that have fans that extend under the fin-stack of the HS (NH-D15, a few Thermalright, etc) and AIOs with fans on the block (ARCTIC Liquid Freezer series, a couple of others).

[-]

imaginary_num6er@reddit (OP)

More like their entire business model was on the Asetek patent that expired this May

[-]

NeedhelpfromYOU@reddit

Thank you Thermalright ❤️