Corsair H70 Internal Testing Results

[roman59]

Not had a chance to do a burn test yet, want to let the stock paste settle down etc, only fitted it around 3 hours ago :)

 

Yes I am running 2 Vipers on it which were on my old H50, at presEnt temps seem about the same at idle, we will see what they are like under load.

 

Cheers

[andyvee]

ok will look forward to it as you have done exactly the same as i would , replace only h70 and keep existing fans etc , stock paste does'nt require burning in imo , never noticed any diff in h50 .

Your idle seems to match what the reviews have got h70 v h50 push/pull so far ( very underwhelming ) but will see what you get under load compared to your old h50 temps .

[roman59]

Forgot to mention because the case pulls in air at the bottom & pushes it out of the top the H70 is pulling air out of the case & not pulling it in.

 

Didnt want to mess up the airflow over the motherboard.

 

Cheers

[everettfsj]

Well here's another "not so glowing" H70 review from Overclock3D;

 

http://www.overclock3d.net/reviews/cases_cooling/corsair_h70_review/3

 

The last chart shows the H50 (one intake fan only) beating the H70 (two fan push/pull) by 1C for a 3.6GHz OC full load on an i7-930 CPU. :eek:

 

See page 4 where they give the H70 a bronze award.

[Yellowbeard]

Well here's another "not so glowing" H70 review from Overclock3D;

 

http://www.overclock3d.net/reviews/cases_cooling/corsair_h70_review/3

 

The last chart shows the H50 (one intake fan only) beating the H70 (two fan push/pull) by 1C for a 3.6GHz OC full load on an i7-930 CPU. :eek:

 

See page 4 where they give the H70 a bronze award.

 

It's odd to me that you would question our testing methodology and the competence of our staff (yes, some of them are experienced engineers just like you) yet throw this review out as proof of anything. Again, this review is clearly not an apples v apples comparison to any of our internal work.

[andyvee]

Well here's another "not so glowing" H70 review from Overclock3D;

 

http://www.overclock3d.net/reviews/cases_cooling/corsair_h70_review/3

 

The last chart shows the H50 (one intake fan only) beating the H70 (two fan push/pull) by 1C for a 3.6GHz OC full load on an i7-930 CPU. :eek:

 

See page 4 where they give the H70 a bronze award.

 

If you watch the video for this , the testing appears to being performed on an open platform with no case at all which would explain why the dh-14 was ( and always is with their reviews :roll: ) the best cooler but in the hexus review the dh-14 was beaten by the intel ref cooler most of the time , all down to testing environment and controls ( or lack of ) .

[everettfsj]

In the very same review your quoting it tells you that the pump is 30mm :p:

 

 

Thats what comes in the box so reasonable they would test like that , those of us with h50 p/p will likely be using completely different fans as well , we'll all have our preferred fans and if buying the h70 would replace the fans with some high static pressure ones anyhow with that thickness of rad .

 

 

Firstly see above , also the Hexus review while good was hardly demanding , they ran prime for 30 minutes , a full stability test ~ 8 hours would show the extra cooling reserve of the h70 , not that i would expect Hexus to do that in a review but users like me do and will when the product is available .

 

 

I would'nt presume that Corsairs target market is h50 users more likely another company's air/liquid cooling . Can't see anyone complaining as the reviews are out there showing the differences although case and position in the home will be a big variable for people , don't know anyone who has their machine on a table in the middle of a room , more like in a corner against a wall .

 

 

Too many presumptions , i have no idea what people work at either Corsair or Asetek as they have no idea what i do ( involves 6 data centres sipping 12Mw :bigeyes: )

 

 

Thats your choice , from my position i'm waiting for real people reviews once stock gets out , at the moment i'll probably keep my h50 as its performing nicely until its 2 years old and then replace with the h70 or its current replacement but if real people show better than shown improvements i may go for it

 

Well thank you for spotting the 30mm height, my bad. :sigh!:

 

As to the Corsair testing of the H50, until told otherwise by Corsair, I'll just have to assume a single fan (Corsair's) intake for the H50 versus a double fan/higher CFM push/pull for the H70 (e. g. both out-of-the-box default configurations with the H70 fans set on their highest setting (which should have significantly higher in situ CFM then the in situ CFM for the H50 as neither account for any pressure restrictions upstream/downstream).

 

As for Prime95, I've run it hundreds of times, and 30 minutes is about 3X the necessary time needed to reach maximum CPU core temperatures, you only need to run it for much longer time spans for numerical stability purposes only. I have dozens of RealTemp Excel 2010 datasets showing the CPU cores reaching maximum ~6 minutes into the tests (e. g. 99.44% of the final delta asymptotic temperature is good enough for temperature measurement purposes).

 

As to Corsair's target market, why pit the H50 directly against the H70 as their only direct comparative test to date? It becomes quite clear that they are going after current H50 users in addition to new users.

 

I have my computer on a table in the middle of a room. :biggrin:

 

As to questioning my expertise or Corsair's or Asetek's, in my case at least, Google USACE ERDC, than look at their 3rd straight Army best R&D laboratory PR, where you will see something called the LMCS (first picture), I worked on that project for four years, specifically for my world class experimental expertise, and I can give you 50-100 professional peer references from the ERDC to boot.

 

I too might drop ~$110, and do my own experimental testing, there's a bottom line to be had, and I'm the only one capable (from what I've seen here in these forums to date) of doing the necessary controlled experiments with full disclosure of said test results. ::pirate::

[Yellowbeard]

As to questioning my expertise or Corsair's or Asetek's, in my case at least, Google USACE ERDC, than look at their 3rd straight Army best R&D laboratory PR, where you will see something called the LMCS (first picture), I worked on that project for four years, specifically for my world class experimental expertise, and I can give you 50-100 professional peer references from the ERDC to boot.

Thanks but no thanks. I'll take your word for it. In my experience, people that blow their own horn too frequently really aren't "been there n done that" kind of guys. I have found that to be especially true with military and police people I've known over the years. ::pirate::

 

I too might drop ~$110, and do my own experimental testing, there's a bottom line to be had, and I'm the only one capable (from what I've seen here in these forums to date) of doing the necessary controlled experiments with full disclosure of said test results. ::pirate::

LOL...there is no need to be insulting and you really have no clue about whom you are speaking. This is a pretty ridiculous statement seeing as how you cannot possibly know any of the qualifications or experience of the people involved. Just because there is no white paper here, this does not mean our testing was improperly done. If you want to review the product, please do and post your results here.

[SaintGermain]

I'm not a very advanced user, but test the H70 out of a case is it really a good method of testing temperature :confused:

[everettfsj]

It's odd to me that you would question our testing methodology and the competence of our staff (yes, some of them are experienced engineers just like you) yet throw this review out as proof of anything. Again, this review is clearly not an apples v apples comparison to any of our internal work.

 

However, it is a review non the less, is it not?

 

Logically, I don't expect anyone to reproduce your results, because we don't (and won't ever) have full disclosure of your test conditions, now do we?

 

And I don't expect, or want, full disclosure.

 

There's engineering, and then there is marketing, and I am of the opinion that your H50 versus H70 comparison leans more towards the marketing side of things, given the current timing of things (your PR, some smattering of internal test results, than the "independent" reviews (with links to the most favorable reviews on your website no doubt to be seen in the foreseeable future), than concurrent (or slightly lagging) release to the mass market, is how it usually works.

 

I'm channeling that GEICO "Honest Abe" commercial, if you get my drift. ;):

[everettfsj]

Thanks but no thanks. I'll take your word for it. In my experience, people that blow their own horn too frequently really aren't "been there n done that" kind of guys. I have found that to be especially true with military and police people I've known over the years. ::pirate::

 

LOL...there is no need to be insulting and you really have no clue about whom you are speaking. This is a pretty ridiculous statement seeing as how you cannot possibly know any of the qualifications or experience of the people involved. Just because there is no white paper here, this does not mean our testing was improperly done. If you want to review the product, please do and post your results here.

 

I'm a civilian and always have been. :confused:

 

You've presented a data point, it has been dually noted. I have not ever suggested that your test results were improperly done, this form of argument is also known as a straw person argument, and it is one form of an informal logical fallacy. :sigh!:

 

I will post my own test results of the H70 versus the H50 here at some future date TBD.

[andyvee]

Forgot to mention because the case pulls in air at the bottom & pushes it out of the top the H70 is pulling air out of the case & not pulling it in.

 

Didnt want to mess up the airflow over the motherboard.

 

Cheers

 

Should'nt affect the results , all your after is if the h70 is cooler under load in exactly the same setup as the h50 , now get testing :D:

[everettfsj]

Should'nt affect the results , all your after is if the h70 is cooler under load in exactly the same setup as the h50 , now get testing :D:

 

Using the exact same set of fans and settings for testing both units, to boot. :cool:

[everettfsj]

I'm not a very advanced user, but test the H70 out of a case is it really a good method of testing temperature :confused:

 

Actually that's the best way to test any heatsink, as you know the ambient temperature on both sides of the heatsink are the same, and most likely at their lowest possible temperature, that being ambient room temperature.

 

If I had that option, I would always mount the H50/H70 on the outside of my case.

[roman59]

Hey guys

 

I am really happy with the H70, the improvements on the pump & copper baseplate is fantastic, really hard to believe that there is actually a pump inside the casing.

 

The swivel on the 90 degree connections from the pump is a nice touch as the pipes can be moved nicely to stop the tubes from bending & being under a lot of tension.

 

Not had a chance to really test it under load, never had any complaints about the H50 at all, its a brand new product which I was lucky to get hold of as my local supplier is now out of stock.

 

I always ran 2 fans on my old H50 so thats no different with the H70, the H70 pump cant be heard at all, there was a slight hum from the H50 so good news there.

 

My old H50 will be going into my brothers pc so he is happy as well.

 

Corsair is not forcing anyone to buy this product you pay your money and take your choice & I am glad I purchased my H70 & I am a happy customer.

 

Cheers

[SaintGermain]

Thank you for your reply Everttfsj.

I thought when designing hardware, the designer was testing equipment in the configuration that used to reflect actual figures. Corsair maybe performed a test in a Obsidian 800D case for example, not with a Torture Rack.

Regarding the performance of H70 in itself, I understand your skepticism when we see some results but I do not understand what the company Corsair have to win to "transform" the results. When we enjoyed such a reputation, at least in Europe, it is not possible to think that. In my opinion, the test procedure used by Corsair does not correspond to the test procedure of the test sites.

Finally, I would also say that I sometimes got better results on my computer with my H50 that some test sites...and yet I am not an expert::pirate::

[everettfsj]

Hey guys

 

I am really happy with the H70, the improvements on the pump & copper baseplate is fantastic, really hard to believe that there is actually a pump inside the casing.

 

The swivel on the 90 degree connections from the pump is a nice touch as the pipes can be moved nicely to stop the tubes from bending & being under a lot of tension.

 

Not had a chance to really test it under load, never had any complaints about the H50 at all, its a brand new product which I was lucky to get hold of as my local supplier is now out of stock.

 

I always ran 2 fans on my old H50 so thats no different with the H70, the H70 pump cant be heard at all, there was a slight hum from the H50 so good news there.

 

My old H50 will be going into my brothers pc so he is happy as well.

 

Corsair is not forcing anyone to buy this product you pay your money and take your choice & I am glad I purchased my H70 & I am a happy customer.

 

Cheers

 

I much prefer quantitative properties that can be measured, than qualitative properties that can not be measured.

 

My favorite color is green, I think I'll paint my H50 green, that way the humming that wasn't there in the first place, still won't be there.

 

Maybe I should also put some green racing stripes on my case.

 

Than my PC will be really c-o-o-o-o-o-o-o-o-o-o-o-l! :cool:

[everettfsj]

Thank you for your reply Everttfsj.

I thought when designing hardware, the designer was testing equipment in the configuration that used to reflect actual figures. Corsair maybe performed a test in a Obsidian 800D case for example, not with a Torture Rack.

Regarding the performance of H70 in itself, I understand your skepticism when we see some results but I do not understand what the company Corsair have to win to "transform" the results. When we enjoyed such a reputation, at least in Europe, it is not possible to think that. In my opinion, the test procedure used by Corsair does not correspond to the test procedure of the test sites.

Finally, I would also say that I sometimes got better results on my computer with my H50 that some test sites...and yet I am not an expert::pirate::

 

The aim of any air/water cooling system is to get as close to ambient temperatures as is possible.

 

You do an exhaust push/pull, or you do an intake push/pull, or you move fans around, or you change the airflow in a multitude of ways, with that single goal in mind.

 

That goal is: How best to get ambient air temperatures into, and out of, your heatsink.

 

For any heatsink inside a case, that means isolating hot spots, if possible.

 

That means moving as much ambient outside air, into, through, and outside the case as efficiently as possible, to yield the lowest CPU core temperatures, barring user constraints (e. g. noise levels, etceteras).

 

Corsair has, so far given us a total of one data point, and that's it.

 

Obviously, I am not satisfied with one data point, especially when there is a claim put forth as "the worst case scenario possible" or some such verbiage.

 

Heck, I can think of two worst case scenarios; 1) Shut the pump off, and 2) shut the fan(s) off. :roll:

 

How does one go about testing for "the worst case scenario" in the first place?

 

Usually, in experimental design, a range of dependent variables (space, time, and mass being the independent variables, following the laws of geometric, kinematic, and dynamic similitude) are selected through something called dimensional analysis, which I learned way back in 1973, but the basic methodology has been around for quite some time (centuries).

 

One then set's up a test matrix, this can be multi-dimensional, depending on the number and degree of the most important dependent variables.

 

You then run a battery of tests filling out said test matrix (or test matricies).

 

Now you are truly and objectively in some position to state "the worst case scenario" or the "best case scenario" or whatever from your testing results based on some desired properties or system constraints or some minimum or maximum criteria, etceteras.

 

However, your "worst case scenario" or whatever is always bounded by the tested regimes, and that this should be self evident and all too obvious.

 

Finally, define the term "worst case scenario" which in this case is clearly ill defined, for example, if the H50 and the H70 gave the exact same temperature, that would most certainly be a "worst case scenario" for Corsair. :roll:

 

Who would want to pay a price premium for the exact same temperature results? No one would, of course, and Corsair would not release such a result to begin with, again this should be all too obvious.

 

Certainly a worse case would be a higher ambient room temperature, e. g. mine is usually 30C (86F), so obviously that is worse that the "worst case scenario" Corsair has shown to date (25C).

 

I could go on, and on, and on, but I'll stop right here. :sigh!:

[Yellowbeard]

 

Obviously, I am not satisfied with one data point, especially when there is a claim put forth as "the worst case scenario possible" of some such verbiage.

 

Heck, I can think of two worst case scenarios; 1) Shut the pump off, and 2) shut the fan(s) off. :roll:

Since you seem to hung up on semantics or verbiage, let's look at this. No pump or no fans would not be a worst case scenario of "heat". Either or both of these would be the result of improper installation or component failure and not of "heat" which is generically referring to the air temperature in the cooling environment that is increased above the ambient temperature. And, both improper installation and component failure are far outside the scope of the testing and is not reflected in this very simple forum post here.

 

You can dissect what is a simple marketing oriented statement until the cows come home if you like. But it won't change the results.

 

I could go on, and on, and on, but I'll stop right here. :sigh!:

Very cool :sunglasse

[vvilliamm]

where did he get the h70 0.0 im still waiting

[everettfsj]

Since you seem to hung up on semantics or verbiage, let's look at this. No pump or no fans would not be a worst case scenario of "heat". Either or both of these would be the result of improper installation or component failure and not of "heat" which is generically referring to the air temperature in the cooling environment that is increased above the ambient temperature. And, both improper installation and component failure are far outside the scope of the testing and is not reflected in this very simple forum post here.

 

You can dissect what is a simple marketing oriented statement until the cows come home if you like. But it won't change the results.

 

Very cool :sunglasse

 

You did see the emoticon :roll: "Rolls Eyes (Sarcastic)" as in I was not in the least bit being serious.

 

I'm somewhat more serious when I mentioned 30C (86F) as a "worst case scenario" for me at least, because that's my normal room temperature, and that has been one ambient condition I've tested constantly with the H50, as I've noticed a proportional loss in temperature cooling efficiency correlated to rising ambient temperatures.

 

I'm in the deep south (Vicksburg, MS), so I'll have to wait until cooler temps arrive in the fall to test outside where I'll have a much larger temperature range and stable daily highs/lows to play with.

 

Every time I test at 20C inside with the AC blasting away for a few hours to go from 30C to 20C, I have wear four t-shirts (two of them long sleeve) and a polar fleece jacket, three pairs of socks, and long pants.

 

25C is also very cold for me.

 

Even then, I can only stand it for a few hours, as it chills me to the bone.

 

Later,

Frank

[Wired]

where did he get the h70 0.0 im still waiting

 

Who are you referring to?

[Laust]

He must be referring to roman59.

[roman59]

I purchased the H70 from Scan Computers in the Uk.

 

Cheers

[andyvee]

I purchased the H70 from Scan Computers in the Uk.

 

Cheers

 

Just checked them and there out of stock again :(: , ebuyer have a few but won't use them as they always screwed up my orders so i'll wait for Scan to get some more