Corsair Hydro Series H100 CPU Cooler Tested

[RJR]

why are we not testing max overclocks on the coolers? sure it takes some time, but that is what most people buy coolers for (or is that only me?)

Dave will have all the reviewers pushing the clocks higher when the "Ivy's" are out this spring, I'm guessing at least 4.5 Ghz (common oc for SB) as the OC bench speed

[ccokeman]

why are we not testing max overclocks on the coolers? sure it takes some time, but that is what most people buy coolers for (or is that only me?)

 

If you read the bottom of the testing page I have some unofficial testing on the max clock speed I can run prime 95 at. 4.2GHz is it for my chip.

[Bosco]

why are we not testing max overclocks on the coolers? sure it takes some time, but that is what most people buy coolers for (or is that only me?)

 

 

+1 Who buys a $119 cooler to overclock an i7 920 to 3.320GHz?

 

When you consider the amount of wear and tear we all put on our systems you will understand.

 

I don't need need guy's pushing max for every cooler and then something blowing up. Not every cooler can hold temps low enough to go that high.

 

Every motherboard is different, as soon as one board overclocks higher then another people are going to start asking why we don't use that board. Shakes head no matter what you do or how you do it people will always complain, the way we do is the way I have choosen to be effective on time and hardware.

 

However you will see the odd review where things are maxed out but don't expect it everytime. But Frank is pretty good about showing max on most of his stuff to give you guys an idea.

 

Starting on the next set of CPU's you will see a higher starting point of overclocks when we do testing, as to what the speed will be I do not know yet.

[El_Capitan]

I understand... but even if it's not compared to others, one idea would be to keep overclocking to see when it will hit a certain temp (say 72C). Not every cooler can cool a hot overclocked chip, but every cooler can have a threshold overclock limit with temperatures being a factor. It's one thing that no overclocking review source does, but I think it's a good test.

[Waco]

It's one thing that no overclocking review source does, but I think it's a good test.

It's also one thing that depends entirely on ambient temps, the chip in question, and HUGE amounts of time to get the chip to hit the specified maximum temperature.

 

When you also bring in the fact that no two chips are the same and even at the same settings they will run different temperatures your suggestions are impossible.

 

 

I may not be a reviewer but I know the crap they go through to get you the numbers you want. Would you risk your hardware every single day, combined with hours of testing every single day, just to completely characterize a cooler you got for review?

 

I know I wouldn't...and I appreciate all that the reviewers do to get the information we need out in the open.

[El_Capitan]

It's also one thing that depends entirely on ambient temps, the chip in question, and HUGE amounts of time to get the chip to hit the specified maximum temperature.

 

When you also bring in the fact that no two chips are the same and even at the same settings they will run different temperatures your suggestions are impossible.

 

 

I may not be a reviewer but I know the crap they go through to get you the numbers you want. Would you risk your hardware every single day, combined with hours of testing every single day, just to completely characterize a cooler you got for review?

 

I know I wouldn't...and I appreciate all that the reviewers do to get the information we need out in the open.

So, I don't understand what you're saying. First, the most important, is what you're saying that they test with different chips? Second, are they testing with the same ambient temps for all the coolers?

 

I'm not a reviewer either, and yes, I know what they go through to get the numbers, not the numbers I want, but if they didn't like it, they don't have to be a reviewer.

 

Let me make this easy on how to test.

 

1. The ambient temperature must be a temperature of no greater than 72C or less than 67C.

2. The same motherboard and CPU and thermal compound must be applied.

3. CPU Cooler #1 gets tested at stock, moderate overclock, and a continued overclock until the highest core temp reaches 72C.

4. CPU Cooler #2... etc.

 

That simple, end of story.

[Bosco]

Do you have any idea what Mike and Frank go through to make that ambient temperature is the same every single time.

 

Poor Frank has to shut off rigs, adjust the heat or the AC just to make sure the temps are bang on every time.

 

As for your points:

 

3. CPU Cooler #1 gets tested at stock, moderate overclock, and a continued overclock until the highest core temp reaches 72C.

 

That will not work, there is to many bad chips out there that need extra voltage that causes it to reach higher temps way to early but still overclock just as well as a chip that doesn't need voltage. The biggest issue is every damn chip is different and there is nothing can do about it. We pick a overclock speed and stick with it. We have also see where chips start to produce higher temps after use, Frank has changed CPU's a few times over the years as have I.

 

We just don't have the time to cater to everyones wants and needs I would love to cover everything but we just can't do it.

[Waco]

So, I don't understand what you're saying. First, the most important, is what you're saying that they test with different chips? Second, are they testing with the same ambient temps for all the coolers?

 

I'm not a reviewer either, and yes, I know what they go through to get the numbers, not the numbers I want, but if they didn't like it, they don't have to be a reviewer.

 

Let me make this easy on how to test.

 

1. The ambient temperature must be a temperature of no greater than 72C or less than 67C.

2. The same motherboard and CPU and thermal compound must be applied.

3. CPU Cooler #1 gets tested at stock, moderate overclock, and a continued overclock until the highest core temp reaches 72C.

4. CPU Cooler #2... etc.

 

That simple, end of story.

Wow.

 

There are many reviewers. All of them have different chips - thus all of them will get different results when overclocking. Also, each of them have different ambient temps.

 

So let me explain why your "easy setup" doesn't work:

 

1. The ambient temperatures vary - it's a complete PAIN IN THE ARSE to get them similar between tests. I have no idea what you're referring to when talking about an ambient temperature between 67 C and 72 C...those temps are WAY past anything on this planet...

2. The same motherboard and CPU and thermal compound cannot be maintained between every test between every reviewer on EVER day. How often do you set up a cooler and get the EXACT same spread of thermal paste?

3. This depends completely upon the specific CPU, motherboard, and ambient temperature...

4. Yeah, see #3.

 

 

It's easy to bitch about how things are done until you realize just how much work goes into delivering a review with just stock speeds and temperatures. What you're asking for is insanely difficult.

 

 

 

 

EDIT: Bosco -

Edited September 17, 2011 by Waco

[Bosco]

Just let me clarify something here.

 

CPU cooling is only done by usually 2 main guys thats Frank and Mike. However they do not share the results, thats why when you guys start complaining about certain cooler missing from each review hence the other guy tested it. This is the same situation when comes to cases as well.

 

When Frank writes a cooler review all of those numbers are his, tested in his house with his test bench, along with the yelling wife when Frank turns off the heat to get the temps down. ( I Love those day's)

 

When Mike writes his reviews samething all of those are his and nobody elses.

 

Just wanted to make sure that was cleared up.

[El_Capitan]

Do you have any idea what Mike and Frank go through to make that ambient temperature is the same every single time.

 

Poor Frank has to shut off rigs, adjust the heat or the AC just to make sure the temps are bang on every time.

I'm sorry, but I live with my gf who thinks it's always too cold when I always think it's too hot. The only way I can get the temperature right is to open a few windows with the AC on so she doesn't get cold. I know what that's like. It's a pain, but it has to be done, and eventually it gets easy to do.

 

As for your points:

 

3. CPU Cooler #1 gets tested at stock, moderate overclock, and a continued overclock until the highest core temp reaches 72C.

 

That will not work, there is to many bad chips out there that need extra voltage that causes it to reach higher temps way to early but still overclock just as well as a chip that doesn't need voltage. The biggest issue is every damn chip is different and there is nothing can do about it. We pick a overclock speed and stick with it. We have also see where chips start to produce higher temps after use, Frank has changed CPU's a few times over the years as have I.

That's fine. The only thing you have to do is stick with the lowest common denominator. So if 3 reviewers have the same chips, you pick the least efficient chip and the other two chips run at those same high voltages, even if they can overclock to the same frequency at lower voltages. How is that difficult?

 

We just don't have the time to cater to everyones wants and needs I would love to cover everything but we just can't do it.

You don't have to cater to everyone's needs, just what the market is intended for it's users. For example, having a $119 self contained liquid cooling unit running at 53C with an i7 running at 3.32GHz when a $25 Cooler Master Hyper 212+ with an i7 920 at 3.42GHz runs at 56.1C (http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/19383-cooler-master-hyper-212-plus-cpu-cooler-review-10.html), people are going to get confused and wonder what the value is with the extra $94 and an OCC Gold award rating (the Cooler Master Hyper 212+ got an OCC Silver award).

[Bosco]

I'm sorry, but I live with my gf who thinks it's always too cold when I always think it's too hot. The only way I can get the temperature right is to open a few windows with the AC on so she doesn't get cold. I know what that's like. It's a pain, but it has to be done, and eventually it gets easy to do.

 

 

That's fine. The only thing you have to do is stick with the lowest common denominator. So if 3 reviewers have the same chips, you pick the least efficient chip and the other two chips run at those same high voltages, even if they can overclock to the same frequency at lower voltages. How is that difficult?

 

 

You don't have to cater to everyone's needs, just what the market is intended for it's users. For example, having a $119 self contained liquid cooling unit running at 53C with an i7 running at 3.32GHz when a $30 Cooler Master Hyper 212+ with an i7 920 at 3.42GHz runs at 56.1C (http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/19383-cooler-master-hyper-212-plus-cpu-cooler-review-10.html), people are going to get confused and wonder what the $89 extra value is.

 

Where is becomes difficult is when someone looks at Frank's review and then goes and looks at Mike's review and see's coolers at different prices and look at other review sites and see higher temps. Then we have to explain that its different temps because more voltage was needed to get to a certain overclock. Its a pain in the butt that chips are all different and are some are really way off depending on the stepping.

 

As for your last statement, you are reading more into this, we don't do that. We keep the speeds the same everytime, you are making up numbers to make your point. I understand where you are coming from, but with us using the same speeds all the time it comes down to temps not the overclock.

 

I am not about to start blowing chips and boards to prove how high everything will go every single time. If there is a point to it we will do it, but for basic cooler testing we don't do it. As I said before come the next batch of chips you will see higher overclocks to start off at compared to what we use now.

[El_Capitan]

Well, being the nerd that I am, I always look at all the different reviews for an item at all the different review sites. There are some reviews I don't put much stock into because of their test settings, and some reviews I definitely pay attention to because the test settings and results are closer to what my target configuration will want to resemble. Just about every review at OCC is top notch, but the only reviews I don't put much stock into are the CPU cooling reviews. I'm not intending to be harsh (though I'm sure I'm sounding that way), just wanting to provide constructive criticism on hoping that it will improve.