i was using 3/4in, where the disconnect (10mm id) was a severe hit on flow. i used eheim as there was no one making dedicated LC pumps (that i could get in a local store), quite the opposite: when the impeller was replaced with a ceramic one, they would work great. one of the reason why a good shop will still offer at least a few (different) eheim pumps.

drac:

I'm curious to know how long a liquid cooling setup like this would last, 4 years (or more preferably) no maintenance or parts failing? I'm updating my system soon and was set on getting an air cooler, but now I'm thinking this could be an option. I just want to basically set and forget until my next CPU, apart from 1 or 2 year GPU updates.

chispy:

Sounds good on paper , but good luck with all the flow restrictions that quick disconnect brings ( I highly advised against using them , very restrictive item as per my own experiences ). Absolutely nothing beats a well thought out , well plan , proper water custom loop. ;) http://i64.tinypic.com/35k3f5j.jpg

I agree, I think if you're gonna go with water cooling you've gotta go custom loop to see any major advantages over good air cooling. I've done some comparative testing (over on notebookreview forums in their desktop section), where I compared my air cooled delidded CPU (Noctua NH-D14) vs another guys AIO 360mm radiator cooled delidded CPU - and at the same wattage load on the CPU my air cooled CPU provided the same temperatures. I was surprised to see this as my fans on my air cooler are running at minimum speed (silent), so I can gain another 5 degC drop from increasing the fans, so in that comparative test I did, my air cooler was actually more efficient than his AIO 360mm radiator cooler. I have to say I spent many hours optimising the airflow in my computer case by experimenting with different fan positions / fan configurations / different combinations of fan RPMs - but if done right air cooling is as good as some of the best AIO liquid cooling, and you also don't have the worries of water leaks and there's less scope for failures, and it's cheaper! My air cooled system is also silent at load too. I think you gotta go custom loop liquid cooling if you want anything done significantly better, and the product in this article looks full of compromises, I wouldn't expect stellar performance from this product.

fry178:

i was using 3/4in, where the disconnect (10mm id) was a severe hit on flow. i used eheim as there was no one making dedicated LC pumps (that i could get in a local store), quite the opposite: when the impeller was replaced with a ceramic one, they would work great. one of the reason why a good shop will still offer at least a few (different) eheim pumps.

That sound like the old poppet valve also know as needle, ball valve unlike the flat valve which has better flow which what the EK system may have

@Robbo9999 there was either something wrong with the aio and/or the setup. an air cooler has to fight the heat that builds up (to a certain degree) inside the case. any radiator/aio setup as exhaust, releases the cpu heat outside the case (something the air coolers will never be able to do), thus gets cooler air going thruy the rad (vs what the air cooler gets from sitting inside the case). right im using a H60i with a low pressure 16dB fan (vs stock HP Corsair fan), and im running temps very close to my custom loop (2-3*C), that had way better cpu block, way more powerful pump and larger hoses (3/4in ID) an external reservoir/radiator ,while ambient room is now about 1*C higher.

fry178:

@Robbo9999 there was either something wrong with the aio and/or the setup. an air cooler has to fight the heat that builds up (to a certain degree) inside the case. any radiator/aio setup as exhaust, releases the cpu heat outside the case (something the air coolers will never be able to do), thus gets cooler air going thruy the rad (vs what the air cooler gets from sitting inside the case). right im using a H60i with a low pressure 16dB fan (vs stock HP Corsair fan), and im running temps very close to my custom loop (2-3*C), that had way better cpu block, way more powerful pump and larger hoses (3/4in ID) an external reservoir/radiator ,while ambient room is now about 1*C higher.

Interesting points, but with air cooling CPU the heat doesn't have to build up in the case at all if it's done right - and in my case the case air temperatures don't increase at all over a 1 hour Prime95 run - measured by System temperature from motherboard and GPU temperature (GPU fan inactive). The reason for the zero increase in temperature during CPU stress is that the Noctua NH-D14 air cooler is in an exact direct line with the rear exhaust fan (vent grill also cut out to increase exhaust flow & lower noise). The exhaust fan is only about 2cm away from the exhaust side of the air cooler, so the case exhaust fan actually helps air pass through the cooler - all the air that exits the back of the cooler immediately leaves the case. I also have a 140mm exhaust fan sat above the CPU cooler in the top panel - hot air from the CPU has zero chance of moving elsewhere in the case. The air cooler is also quite close to the top panel, which is ventilated, so it can pull a good portion of air straight from outside of the case. There is really no build up of hot air from CPU stress. There is a slight build up of hot air in the case from GPU stress though, 100% GPU stress on looped Graphics Test 1 of Firestrike Extreme raises case air temperatures by about 5 degC over idle, but again the CPU is not affected by this whole 5 degC temperature rise as a good portion of it's air is coming directly from the ventilated top panel. I don't think there's anything wrong with the other guys AIO I was comparing against. He's a very technology experienced and technical guy who is well known over on notebookreview forums for modifying laptops, overclocking them, and setting records with his overclock attempts - so he understands hardware & temperature control. I'll show you a pic of my CPU temperatures after a 20min run of OCCT, can't upload them here, so here's link to where I did the testing (post #204 at the following link, on that page): http://forum.notebookreview.com/threads/official-nbr-desktop-overclockers-lounge-laptop-owners-welcome-too.810490/page-21 EDIT: With regard to your Corsair H60 AIO, it's not one of the best AIO, so I'm surprised to hear it's only a couple of degC off your custom water loop - sounds like there's something wrong with your custom water loop. If you look at Guru3D reviews of the H60 you can see it massively underperforms the H110 (http://www.guru3d.com/articles_pages/corsair_h90_review,12.html), while my NH-D14 air cooler performs as well as an H110 according to guru3d (http://www.guru3d.com/articles_pages/cooler_master_nepton_240m_review,11.html). So good air cooling does kick some ass! Seems like there was something wrong with your custom water loop if it didn't really outperform your H60 AIO.

Didnt mean chipset etc, but the temp inside the case, that the cooler breathes, and i didnt mention im mainly interested in SILENT cooling.. Unless you have a high amount of airflow AND an exhaust fan right next to the cpu heatsink, the cooling perf shifts towards lc. Most cases dont have the rear fans that close, that there is no air recirculating (and getting reused to cool the cpu). And once its about running the rig silent, which im interested in, the heat from the cpu ends up mainly inside the case. Even when i had a case with one 140 fan right behind the cpu cooler and 2 above it... And the almost indefinite setups you can have when it comes cooling, "one" chart doesn't tell the whole story, as i have had the last 5 setups with an aio vs air (large heatsink with four 8mm heatpipes and two 140mm fans), and at inaudible noise levels, dumping the heat outside (lc) does lower temps of gpu/cpu by around 20-30*C depending on load and gpu used. Just the fact that hilbert tested the hydro without a case, is sub-optimal for me, as you have different airflow for the rad vs installed in a case, so the numbers are good indicators, but a bit off from real life (except maybe for ppl using a test bench)..

fry178:

Didnt mean chipset etc, but the temp inside the case, that the cooler breathes, and i didnt mention im mainly interested in SILENT cooling.. Unless you have a high amount of airflow AND an exhaust fan right next to the cpu heatsink, the cooling perf shifts towards lc. Most cases dont have the rear fans that close, that there is no air recirculating (and getting reused to cool the cpu). And once its about running the rig silent, which im interested in, the heat from the cpu ends up mainly inside the case. Even when i had a case with one 140 fan right behind the cpu cooler and 2 above it... And the almost indefinite setups you can have when it comes cooling, "one" chart doesn't tell the whole story, as i have had the last 5 setups with an aio vs air (large heatsink with four 8mm heatpipes and two 140mm fans), and at inaudible noise levels, dumping the heat outside (lc) does lower temps of gpu/cpu by around 20-30*C depending on load and gpu used. Just the fact that hilbert tested the hydro without a case, is sub-optimal for me, as you have different airflow for the rad vs installed in a case, so the numbers are good indicators, but a bit off from real life (except maybe for ppl using a test bench)..

"System" temperature as measured by the motherboard is supposed to indicate air temperature in the case, that value doesn't change during 100% CPU load, nor does the GPU temperature increase at all (GPU fans inactive) during the same 100% CPU load runs - the temperatures inside the case don't increase at all with 100% CPU load, and it's silent too (using the ULV low RPM adapter for the Noctua NH-D14 air cooler). Chipset Temperatures that you mentioned are a different sensor on my motherboard, not the "System" temperature that I'm referring to. I think the Guru3d cooling performance temperature charts that are included in their reviews are very valid and an excellent means of comparison. NH-D14 shown to outperform many AIO closed loop liquid coolers, including your Corsair H60 by a country mile, NH-D14 showed same performance as some of the best AIO liquid coolers such as Corsair H110. I'm still of the opinion given the evidence (both through personal testing, and through reading guru3d reviews) that AIO liquid coolers aren't really an improvement over the best air cooling, and that to get significantly better you have to go custom loop liquid cooling.

@Robbo9999 sure do the numbers show the better cooling, when done the same way (open tray etc). as soon as i run my rig that its completely inaudible (silent freak), air coolers dump the heat inside the case as the exhaust fans dont expel the air "quick" enough. i like to tinker when doing a new setup to get the most out it, and tend to do different fan/cooling setups (each for a few weeks) before making the final setup. and every time the air coolers (only speaking for my ultra silent setups) are worse than the cheapest aio running the rad as exhaust. monitoring the gpu temps also shows i get higher numbers. not saying all those testers/reviews are wrong, but they will never cover EVERY way i can use coolers/fans, especially with all those different coolers/fans and cases. so yes, usually with good airflow, air coolers can match aio, just not in my/every case.. oh and i remembered yesterday, why the temps where almost the same on the custom loop: i was running the 3770@1.36v xD (vs stock now)..

fry178:

i like to tinker when doing a new setup to get the most out it, and tend to do different fan/cooling setups (each for a few weeks) before making the final setup.

Yep, me too. 🙂

fry178:

not saying all those testers/reviews are wrong, but they will never cover EVERY way i can use coolers/fans, especially with all those different coolers/fans and cases. so yes, usually with good airflow, air coolers can match aio, just not in my/every case..

Sure, how a system is setup in terms of case/layout/fans is always gonna influence the performance of a CPU cooler (whether AIO liquid or air) by a large amount.