The Neophyte's Custom Liquid Cooling Guide: How To, Why To, What To Expect
by Dustin Sklavos on September 30, 2013 12:01 AM ESTIt stands to reason that the heat barriers I was running into under air would become less of an issue when I moved to water. That's...partially true. In a weird way, I feel like this review would ultimately have been a lot less useful if things had gone smoothly instead of turning out the way that they had. Early on I mentioned that overclocking is and will always be a bit of a lottery. The halcyon days of Sandy Bridge are behind us and for many of us even Ivy Bridge is going to be a dream of better days.
I used my overclock under air as a fairly appropriate starting point. Since 4.4GHz and 4.5GHz were outside the realm of possibility under air due to thermals, I made the jump to 4.4GHz, and it was here where I ran into two complications: one specific to Haswell and one broad enough that I should've seen it coming even as more of a hobbyist overclocker.
First, when overclocking the CPU, it's wise to set the RAM to the spec of the CPU to take it out of the equation; in the case of the Core i7-4770K, that's 1333MHz/1600MHz. After finding an overclock, though, I neglected to test it under the G.Skill Trident X RAM's rated speed, and that complicates things with Haswell. An overclock stable at 4.4GHz at DDR3-1333 wound up needing even more voltage for DDR3-2133. Haswell's IMC can be particular about memory speed when the cores are heavily overclocked.
The other more broad complication was that I hadn't accounted for the minor voltage bump that 4.3GHz required, or the major one that 4.5GHz did under air. Chips tend to have an inflection point; you can get up to a certain speed at stock voltage or with a slight bump, but once you get there, any increases in speed require substantially boosted voltage. As it turned out, 4.5GHz wasn't happening, and 4.4GHz required a fairly healthy one in and of itself.
4.4GHz at DDR3-2133 and 1.35V ran benchmarks fine, but stress testing for power consumption and thermals with the GTX 780s being pushed at the same time caused its core temperatures to shoot over 90C. I had to run all the fans on the radiators at maximum to cool that 4.4GHz as well. Ultimately a really good overclock just wasn't in the cards for this i7-4770K.
Meanwhile, the two 780s had also apparently hit their wall on air. Watercooling stabilized boost clocks, but even after overvolting in Precision X, I was only able to get them to run at a consistent ~1160MHz on the cores and 7.1GHz on the GDDR5. Temperatures remained incredibly low; a modified BIOS would probably help but is out of the scope of this article. For now, I have to rely on the improved boost stability and 100MHz jump on the CPU to improve performance.
Facebook
Twitter
RSS
106 Comments
View All Comments
hot120 - Monday, September 30, 2013 - link
Awesome article!blanarahul - Monday, September 30, 2013 - link
Hmm.. Can you try cooling those 780s alone? Overclocking the CPU seems pointless on Haswell.valkyrie743 - Monday, September 30, 2013 - link
overclocking haswell is not pointless. just is a pain (same with ivy bridge) cause intel decided to be cheap and not solder the IHS to the cpu. if you do a mild overclock its fine give or take how bad the tim on the cpu/ihs is. but if you plan on doing high overclocks and water cooling like this. you might as well de-lid the cpu and apply your own tim. temps on air (if done right) drop a good 15 to 20C under load. I've seen people hitting 90 C and go down to 70 or less underload. and thats on air.the reason why i have no upgraded from my sandy bridge 2600K. @4.5ghz right now at 1.28 volts and my max temp running intel burn test was 70C (air)
The Von Matrices - Monday, September 30, 2013 - link
Please read my post in response to NeatOman. The result is correct but the reasoning is incorrect.gandergray - Tuesday, October 1, 2013 - link
For information about removing the cpu lid or integrated heat spreader, see the work performed by Idontcare: http://forums.anandtech.com/showthread.php?t=22618... .iTzSnypah - Monday, September 30, 2013 - link
You are cooling way too much with only 600mm worth of radiators and your deltaT is obscene. Take out 1x GTX780 and retest if possible.NeatOman - Monday, September 30, 2013 - link
I think the thermal paste between the cpu and the lid are the limiting factor here, i believe that not only will 4770K do better with better thermal paste in between the lid and cpu on just air cooling alone but also might have a larger difference between the air and water cooling.And of course there is also a full delid which i think wont be much of a threat because with water cooling you don't need the motherboard to support a large heavy cooler.
NeatOman - Monday, September 30, 2013 - link
Sorry, i meant that you wont need to put a lot of pressure like if you where supporting a large air cooler with the motherboard.The Von Matrices - Monday, September 30, 2013 - link
The issue is not the composition of the thermal paste between the die and the lid; it is the thickness of the thermal paste between the die and the lid. It's widely reported that in Ivy Bridge and Haswell there is way too much of a gap between the die and the lid due to the thickness of the glue used to secure the lid to the package. You can solve this by removing the lid, using a razor blade to remove all the glue, then put on new TIM and place the lid back on the package. No matter what new TIM you use you will get drastically reduced temperatures.Either way, Haswell runs hot due to its FIVR, and there's nothing that can be done through beefier heatsinks, delidding, or changing thermal paste that will make it cooler than an equivalently modified Ivy Bridge.
dragosmp - Monday, September 30, 2013 - link
Still, it is incomplete. The thermal transfer formula is simply Rth=rho*L/S, more thermal resistance (Rth) more the temperature delta is high between the source and ambient: deltaT=Power*RthAsuming the power is constant, to decrease deltaT you need to decrease the thermal resistance, so:
*S is the die surface, can't change that
*L is the thickness of paste - you're right, it needs to be as thin as possible; put 2x too much and you have twice the deltaT
*rho - thermal resistivity (1/lambda) - it depends on the material; Intel does use cheap paste with a conductivity around 3; were they to use fluxless solder or at least some AS5 they'd decrease the thermal resitance by a factor of 2 easily, thus offsetting a thicker than needed layer of paste.
My 2 cents: for performance the paste must be removed and replaced with something better plus as you say remove the glue to reduce the thickness. Of course one should be careful not to chip the die, but these two things really help.