The Corsair User Forums  

Go Back   The Corsair User Forums > Corsair Product Discussion > Cooling

Reply
 
Thread Tools Search this Thread Rate Thread Display Modes
  #1  
Old 05-14-2019, 11:11 AM
ltron ltron is offline
Registered User
ltron's PC Specs
 
Join Date: Jun 2015
Posts: 9
POST ID # = 1004550
ltron Reputation: 10
Default H150i Pro Upgrade From Noctua NH-D15 for Zen2

Hi, I currently have a Noctua NH-D15 air cooler and I was thinking of possibly trying my first AIO CLC in the Corsair H150i Pro when I upgrade to an AMD Ryzen 3000 12 core CPU later this year. My NH-D15 will continue to be used in my secondary PC.

My my main reason for switching is space around the CPU socket so removing and installing GPUs is easy (there's virtually no room on my X99 setup with the NH-D15 and it's very hard to remove the GPU), I am also tempted to try something different and go closed loop water. I am not interested in a custom loop due to the maintenance required.

Does the fact that the pump uses a metal impeller mean that the lifespan is longer and chance of failure is significantly lower than with a traditional CLC? Will the performance of the H150i degrade over time due to permeation? If so, by how much?

My priorities are silence, reliability with at the very least a 5 year lifespan if not 10 years, good temperatures, RAM compatibility, being able to transport the PC without removing the cooler and space around the CPU socket.

I don't care much about aesthetics although they are nice to have and the price difference isn't very important to me.

Is this a bad idea? Has anyone else changed from big air to an H150i pro or equivalent?

What is the default fan and the default pump profile for the H150i if Link/iCue is not used, is it 'balanced' for both fans and pump?

Last edited by ltron; 05-14-2019 at 11:21 AM.
Reply With Quote


  #2  
Old 05-14-2019, 07:15 PM
c-attack c-attack is online now
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 8,254
POST ID # = 1004610
c-attack c-attack Reputation: 103
Default

I would not read too much into marketing language for pump and cooler features. There are only a few OEM parts suppliers and design are similar. Differentiation is going to be in aesthetics, fan types, and warranty coverage. Corsair offers a 5 year warranty on most coolers.

I can take an AIO on and off in less than 5 minutes. It might take hours to put on big air tower, especially since you often need to remove the entire motherboard from the case. That check box is an easy yes.

Hopefully no permeation. That would be rather bad. All AIO and custom loops will suffer some degree of evaporation in time. It almost never gets to the point where fluid volume affects performance and I can probably count the number of "low fluid" complaints from last 5 years on one hand. Either way, the warranty is 5 years. At that point, you likely no longer have a use for that specific cooler and are ready for something else in a different shape or size. Every once in a while someone will pop up with a "I've been using this cooler for 7 years story", but most people will move on before that point. Mechanically wearing out the cooler is unlikely. Most reported issues revolve around something else. However, these things are $100-150 product. It is not a piece of furniture and a 10 year useful lifespan it not a realistic expectation. I would not expect my $200 D5 pump to still be pristine 10 years on either.

Tough to say what to expect for the 3000 series. I have seen 135W as potential target for the 8 core. I'll guess the 12 core will be specified at 165-175W, with a real output of 205-225W at full blast. I would say you do want a 360mm. At that level of wattage, you are getting into the area where 280 and 240mm coolers need high fan speeds to keep up. The larger surface area of a 360mm cooler should allow you to keep medium to low fans speeds without substantial temperature penalty. There is no air cooler I would recommend for a 12 core, 7 nm or not. You will be able to save a custom fan curve and one of the three pump speed settings to the cooler hardware, if for some reason you don't run iCUE. However, you would also be giving up your monitoring ability for the coolant temperature.

Keep in mind the big physical difference between a water cooled system and an air tower is the pump. There obviously isn't one in an air tower and adding a small mechanical-electrical device adds noise. Typically water cooled radiators are denser than the big box air tower and this will make the air passing through sound different, typically a bit sharper. This gives better heat dissipation, but with a different sound. I would never try and convince someone a water cooler will be less noisy than an air tower at idle. Where it might be better is at load noise and certainly in terms of load temperatures. You will never have the same heat holding capacity with an air tower that you get with even a small water cooled system. Unlike the air tower, you don't need instant and aggressive fan response to get rid of the heat. With water cooling and especially on a 360mm, you can let the fans stroll along and the system will push out the trash heat in due coarse without substantial negative impact on the CPU temperature.

Last edited by c-attack; 05-14-2019 at 07:18 PM.
Reply With Quote


  #3  
Old 05-15-2019, 07:54 AM
ltron ltron is offline
Registered User
ltron's PC Specs
 
Join Date: Jun 2015
Posts: 9
POST ID # = 1004642
ltron Reputation: 10
Default

Quote:
Originally Posted by c-attack View Post
I would not read too much into marketing language for pump and cooler features. There are only a few OEM parts suppliers and design are similar. Differentiation is going to be in aesthetics, fan types, and warranty coverage. Corsair offers a 5 year warranty on most coolers.

I can take an AIO on and off in less than 5 minutes. It might take hours to put on big air tower, especially since you often need to remove the entire motherboard from the case. That check box is an easy yes.

Hopefully no permeation. That would be rather bad. All AIO and custom loops will suffer some degree of evaporation in time. It almost never gets to the point where fluid volume affects performance and I can probably count the number of "low fluid" complaints from last 5 years on one hand. Either way, the warranty is 5 years. At that point, you likely no longer have a use for that specific cooler and are ready for something else in a different shape or size. Every once in a while someone will pop up with a "I've been using this cooler for 7 years story", but most people will move on before that point. Mechanically wearing out the cooler is unlikely. Most reported issues revolve around something else. However, these things are $100-150 product. It is not a piece of furniture and a 10 year useful lifespan it not a realistic expectation. I would not expect my $200 D5 pump to still be pristine 10 years on either.

Tough to say what to expect for the 3000 series. I have seen 135W as potential target for the 8 core. I'll guess the 12 core will be specified at 165-175W, with a real output of 205-225W at full blast. I would say you do want a 360mm. At that level of wattage, you are getting into the area where 280 and 240mm coolers need high fan speeds to keep up. The larger surface area of a 360mm cooler should allow you to keep medium to low fans speeds without substantial temperature penalty. There is no air cooler I would recommend for a 12 core, 7 nm or not. You will be able to save a custom fan curve and one of the three pump speed settings to the cooler hardware, if for some reason you don't run iCUE. However, you would also be giving up your monitoring ability for the coolant temperature.

Keep in mind the big physical difference between a water cooled system and an air tower is the pump. There obviously isn't one in an air tower and adding a small mechanical-electrical device adds noise. Typically water cooled radiators are denser than the big box air tower and this will make the air passing through sound different, typically a bit sharper. This gives better heat dissipation, but with a different sound. I would never try and convince someone a water cooler will be less noisy than an air tower at idle. Where it might be better is at load noise and certainly in terms of load temperatures. You will never have the same heat holding capacity with an air tower that you get with even a small water cooled system. Unlike the air tower, you don't need instant and aggressive fan response to get rid of the heat. With water cooling and especially on a 360mm, you can let the fans stroll along and the system will push out the trash heat in due coarse without substantial negative impact on the CPU temperature.
Hi c-attack,
thanks for the detailed reply, it's very helpful. The reason I mention the pump and its metal impeller is that Gamersnexus did a tear-down on the H150/H115i/H100i pro coolers that use the new gen 6 Asetek pump (the only coolers I know of that use it) and found that the impeller used is metal rather than the traditional plastic. Given the performance of the new pump is marginally worse than the old one I think they probably did it for reliability reasons.

https://www.gamersnexus.net/hwreview...2-clc-280-more

I very much like the fact that the fans won't keep ramping up in response to any small change in temperature unlike my air cooler and of course ease of install, low noise and space around the CPU socket are big plusses. I'll see how power hungry these new Ryzen 3000 CPUs are before making my decision.
Reply With Quote


  #4  
Old 05-15-2019, 09:01 AM
c-attack c-attack is online now
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 8,254
POST ID # = 1004647
c-attack c-attack Reputation: 103
Default

Quote:
Originally Posted by ltron View Post
Given the performance of the new pump is marginally worse than the old one I think they probably did it for reliability reasons.
I have watched that video a few times before, but I don't remember the conclusion is Gen 6 is worse performing than Gen 5. Either way, I'll tell you right now it is irrelevant to the total system cooling. This has been a hot topic lately, but on both Gen 5 and Gen 6 pump coolers, toggling between the medium speed (~2000 rpm) and highest (~2800-3000) typically has zero difference in coolant temperature and thus zero difference in end CPU temperature. Sometimes someone will report a 1C difference, but then if you want to round up the tenths place... The closest and most recent analysis has been with the 9900K, something likely to be similar in wattage to the 12 core Ryzen 3. Again less than 0-1.0C difference at maximal load for an extended period across many people. This is expected. Even in my custom loop with 450-500W running through it, I get no measurable difference in coolant temperature between 2000 and 4000 rpm on my D5. 4000 is very good for flushing out bubbles after I fill the loop, but aside from that it is just loud.

The reasons for this are fairly straightforward. Most AIO loops are short, have only one component, and thus offer low fluid restriction. You do not need to extreme pump speed or head pressure to create the minimum required flow rate. I took great pains to ensure my custom loop was smooth as well and the results are similar. Extreme Rigs did some heavy radaitor testing a few years back with lots of data at various flow rates. They always come back within tenths of a degree and most testers no longer bother with multiple flow rates because it does not matter, except for the a very low one.
The Pro series gives a good example of what happens when you drop below the minimum pressure level needed. I do not know what that figure is. Likely somewhere below 0.5 gallons per minute and there is not a good way to assess it without cutting it up and measuring. Either way, if you use the 1100 rpm pump speed, there is a definite loss in performance. On the other hand, there will be no difference in coolant temp between the medium and high speeds. So the general advice is to park it on the medium and leave it. Pump speed or flow rate is not a strong determining factor on end CPU temp.

As for the metal vs plastic stuff, I have no idea. Asetek makes tens of thousands of these things. I am sure they have the data to justify the change. I trust them to do what is their best interest. I do know I would never, ever, trade my gen 6 H115i Pro for the model it replaced. The Pro series has been very reliable with an incredible small number of complaints.


This is a typical 360mm radiator and I picked one that is similar in thickness to the H150. Ignore the purple bars since that is one set of fans and different than the rest. The 0.5 G/min bar is more than the 1100 rpm pump speed, but likely less than the balanced 2160 on the Pro series. So the two bars you are looking at are red and green to compare medium to high pump speeds.

Last edited by c-attack; 05-15-2019 at 09:07 AM.
Reply With Quote


  #5  
Old 05-15-2019, 10:03 AM
ltron ltron is offline
Registered User
ltron's PC Specs
 
Join Date: Jun 2015
Posts: 9
POST ID # = 1004658
ltron Reputation: 10
Default

Quote:
Originally Posted by c-attack View Post
I have watched that video a few times before, but I don't remember the conclusion is Gen 6 is worse performing than Gen 5. Either way, I'll tell you right now it is irrelevant to the total system cooling. This has been a hot topic lately, but on both Gen 5 and Gen 6 pump coolers, toggling between the medium speed (~2000 rpm) and highest (~2800-3000) typically has zero difference in coolant temperature and thus zero difference in end CPU temperature. Sometimes someone will report a 1C difference, but then if you want to round up the tenths place... The closest and most recent analysis has been with the 9900K, something likely to be similar in wattage to the 12 core Ryzen 3. Again less than 0-1.0C difference at maximal load for an extended period across many people. This is expected. Even in my custom loop with 450-500W running through it, I get no measurable difference in coolant temperature between 2000 and 4000 rpm on my D5. 4000 is very good for flushing out bubbles after I fill the loop, but aside from that it is just loud.

The reasons for this are fairly straightforward. Most AIO loops are short, have only one component, and thus offer low fluid restriction. You do not need to extreme pump speed or head pressure to create the minimum required flow rate. I took great pains to ensure my custom loop was smooth as well and the results are similar. Extreme Rigs did some heavy radaitor testing a few years back with lots of data at various flow rates. They always come back within tenths of a degree and most testers no longer bother with multiple flow rates because it does not matter, except for the a very low one.
The Pro series gives a good example of what happens when you drop below the minimum pressure level needed. I do not know what that figure is. Likely somewhere below 0.5 gallons per minute and there is not a good way to assess it without cutting it up and measuring. Either way, if you use the 1100 rpm pump speed, there is a definite loss in performance. On the other hand, there will be no difference in coolant temp between the medium and high speeds. So the general advice is to park it on the medium and leave it. Pump speed or flow rate is not a strong determining factor on end CPU temp.

As for the metal vs plastic stuff, I have no idea. Asetek makes tens of thousands of these things. I am sure they have the data to justify the change. I trust them to do what is their best interest. I do know I would never, ever, trade my gen 6 H115i Pro for the model it replaced. The Pro series has been very reliable with an incredible small number of complaints.


This is a typical 360mm radiator and I picked one that is similar in thickness to the H150. Ignore the purple bars since that is one set of fans and different than the rest. The 0.5 G/min bar is more than the 1100 rpm pump speed, but likely less than the balanced 2160 on the Pro series. So the two bars you are looking at are red and green to compare medium to high pump speeds.
That's great info, thanks. I fully agree with you with regard to gen 6 vs gen 5, the performance difference is within the margin of error and not worth giving a second thought. However, the reliability or endurance benefit although not proven beyond all doubt seems very much worth having. I will keep the pump speed on balanced, that's what I intended to do anyway.
Reply With Quote


Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 08:58 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.