Swiftech H220 AIO Review

[grouchon]

the point I'm making is the fuss your making over toms review showing the H100i 4.5c cooler then the H220 yet you have after this Compared to Corsair's H100i: a link that uses 2 different systems to compare the H100i and the H220 showing the H220 10c better then the H100i.

Thank you for clarifying your position; I do believe I already answered similar concerns you posted on OCN as follows:

"I think that you are referring to the CES presentations, and I'd like to clarify that we practically had no other choice but to do it that way because of what we wanted to show: comparing the operating noise and temperature between 4 similarly built systems running at the same frequency in the limited amount of time allowed by the format of this venue. You must realize that we couldn't demo each media representative by assembling disassembling the kits 4 times. It would have taken hours. Of course we knew that the kits would be later on tested in a proper and scientific format. Thus, you can be rest assured that there was absolutely no intention to mislead anyone in any way shape or form "

So the links you are refering to are presentations to the Media.

Here we are discussing a review in proper testing format, and we are only trying to understand why the results so substantially differ from our own. This is scientific due diligence if you will.

 

[grouchon]

It is of course a bit impractical to try and review every unit with pre-applied/provided TIM ...

That Thermal Engineer's opinion sounds dodgy at best btw, I'll wait for actual tests on that rather than accept it now.

Testing:
Yes it is indeed impractical. You only have one shot at testing unit with pre-applied grease. yet this is how users receive it. We understand that different reviewers have different priorities and audiences. The amount and type of data they publish is a reflection of such priorities. Our own goal is to promote user education by providing as much information as possible.

Thermal engineer discussion:
Still waiting for Resident's reponse to my questions.

 

[garih]

welcome to oc3d MR Rouchon, slightly off topic, but it is good to see someone like yourself taking the time to come here and answer questions and generally support your products, nice touch

 

[grouchon]

I would suggest that you employ a better thermal engineer then, one that isn't just a 'yes man'.

See answer below

That Thermal Engineer's opinion sounds dodgy at best btw, I'll wait for actual tests on that rather than accept it now.

see answer below

AMD Phenom II X4 B60
3.9Ghz (modest OC from stock 3.4)
1.342v

No wonder.. your heat load is very modest compared to the 2011 platform.

See our engineer's response below:

Assuming 150-200W heat load (2011 oc/ov)

Rth Difference between 1800 and 2000 RPM: 0.007 C/W to 0.018 C/W depending on the fan / rad.

Taking the 2 extreme cases:
Scenario 1 : 0.007 * 150 = 1.05 C
Scenario 2: 0.018 * 200 = 3.60 C

"Yes-man" still confirms 2C is in the ball park

I hope this clarifies your questions Resident.

welcome to oc3d MR Rouchon, slightly off topic, but it is good to see someone like yourself taking the time to come here and answer questions and generally support your products, nice touch

Thank you, just doing my job. We are very keen in reaching out to the UK community, and the best we know of is to support thru education.

 

[Greenback]

Thank you for clarifying your position; I do believe I already answered similar concerns you posted on OCN as follows:

"I think that you are referring to the CES presentations, and I'd like to clarify that we practically had no other choice but to do it that way because of what we wanted to show: comparing the operating noise and temperature between 4 similarly built systems running at the same frequency in the limited amount of time allowed by the format of this venue. You must realize that we couldn't demo each media representative by assembling disassembling the kits 4 times. It would have taken hours. Of course we knew that the kits would be later on tested in a proper and scientific format. Thus, you can be rest assured that there was absolutely no intention to mislead anyone in any way shape or form "

So the links you are refering to are presentations to the Media.

Here we are discussing a review in proper testing format, and we are only trying to understand why the results so substantially differ from our own. This is scientific due diligence if you will.

And as I put on OCN I not referring to your CES that I understand the limitations, I'm referring to the links you have under the videos and case pictures under H220 Reviews:
Compared to Corsair's H100i:

in actual fact there are 2 h220-h100i comparisons linked in that section where they use different systems and 1 states
we chose Arctic Silver Ceramique 2 as the thermal interface material.

 

[Resident]

No wonder.. your heat load is very modest compared to the 2011 platform.

See our engineer's response below:

Assuming 150-200W heat load (2011 oc/ov)

Rth Difference between 1800 and 2000 RPM: 0.007 C/W to 0.018 C/W depending on the fan / rad.

Taking the 2 extreme cases:
Scenario 1 : 0.007 * 150 = 1.05 C
Scenario 2: 0.018 * 200 = 3.60 C

"Yes-man" still confirms 2C is in the ball park

I hope this clarifies your questions Resident.

Ok, to save continuing the argument, I can see the math works in principle, however in the real world it doesn't quite fit, probably due to the assumptions be slightly out but I'm prepared, if you are, just to put it down to 'one of those bizarre things'

All I care about tbh is that my system is as cool as I want/need it to be.

Also just to echo what Garih said, I commend you taking the time to sign up and post here on OC3D. It's nice to see a CEO come onto the community forums and interact with the very people which potentially consume their products. It's not often that you see that, usually it's 'just' some representative, which is fine, I'm not having a pop at those companies in the least.

So kudos my good man!

 

[Lystfiskern]

We are very keen in reaching out to the UK community, and the best we know of is to support thru education.

What? Im part of the UK community?
And here I was thinking that the interwebz didnt have such boundaries....
So much for posing as a global player

 

[grouchon]

And as I put on OCN I not referring to your CES that I understand the limitations, I'm referring to the links you have under the videos and case pictures under H220 Reviews:
Compared to Corsair's H100i:

in actual fact there are 2 h220-h100i comparisons linked in that section where they use different systems and 1 states
we chose Arctic Silver Ceramique 2 as the thermal interface material.

Hmm.. I was in Shenzhen when this one came out. Let me check it out.

 

[tinytomlogan]

https://docs.google.com/spreadsheet...eHoxckRqazBvM0RTMW5GSFZNRmc&usp=sharing#gid=0


Thats a link to google docs. That is also Linus's aka NCIX Linus test reults. Oh look they are VERY similar if not worse than what I reported...

H100i 71c vs H220 82c?

 

[Feronix]

https://docs.google.com/spreadsheet...eHoxckRqazBvM0RTMW5GSFZNRmc&usp=sharing#gid=0


Thats a link to google docs. That is also Linus's aka NCIX Linus test reults. Oh look they are VERY similar if not worse than what I reported...

H100i 71c vs H220 82c?

Only flaw there (imo) was that he didn't include the results of the H100i with its stock fans

 

[Greenback]

if you go to the kraken review he did you see the h100I with stock fans at balanced 64c.

a lot worse then you showed h220 18c warmer than H100i

 

[jarryy]

as reported in that doc. about 9.7 degree difference between the h220 with NH fans and the h100i (at load) however as the fans in use in that report are $29.95 each at the cheapest here in Australia, so an additional 60 dollars you need to add onto the h220 already is it justified for the 9.7c difference?

 

[tinytomlogan]

Its a shame there is no stock H100i results.

Never the less you can see it is possible to easily get a larger difference than I had.

Similar test kit used, just the fans that vary. Which was all along why the H100i won in my tests and I even said that!

 

[Feronix]

Its a shame there is no stock H100i results.

Never the less you can see it is possible to easily get a larger difference than I had.

Similar test kit used, just the fans that vary. Which was all along why the H100i won in my tests and I even said that!

I'm still considering to buy the H220 though, I just like the looks more (and the pump).

What would you say to convince me to get the H100i besides that it performs slightly better? (I'll be having an i7 3770k on a 24/7 overclock of a very low 4 GHz)

 

[tinytomlogan]

I'm still considering to buy the H220 though, I just like the looks more (and the pump).

What would you say to convince me to get the H100i besides that it performs slightly better? (I'll be having an i7 3770k on a 24/7 overclock of a very low 4 GHz)

Its not for me to convince you either way mate. I tell you what I think in the videos and reviews, beyond that you buy what you like.

 

[Feronix]

Its not for me to convince you either way mate. I tell you what I think in the videos and reviews, beyond that you buy what you like.

Hehe okay, I'll think it over. Guess they both have their pros and cons. Won't have my money for a while anyway

 

[ninogui]

my first post at oc3d..

Because this subject raised enough interest from me to do so.

I kinda like Tom´s reviews. Opposite to most reviewers he charges a lot of emotion onto them so you feel he is really talking from the heart. This is good if you equate them the right way. Most reviewers are noticeably biased towards what they are reviewing and you can feel that if you are in this business for many years (like I am).
However new people should filter both emotional and biased reviews, as well as get the most info you can coming from everywhere before making a judgement on a particular hardware piece. Only then can you average them all and conclude by yourself.

I am almost sure Tom is not biased towards corsair and he would like to have another king of the hill on this (or any other piece of equipment).

Carefully listening to all the H20 review I have no doubts he was really expecting the H20 to perform much better than the H100i. So much it left him getting too much carried away out of emotion. When you go to lengths like waiting for the right time of day to get precise ambient temps and so forth you know he has really tried, but also means something has clearly gone wrong and reason leaves the room.

I do take note of Gabe´s involvement in this and I take my hat to him. This does not happen everyday on almost any other manufacturer. I think Sw are trying to close the gap between all in one and custom builds for a very reasonable price and that´s good

I don´t usually comment on stuff I didn´t test or tried myself, but I felt pulled on this one.

Is the H20 pump louder ? it probably is, I don´t find that surprising given it´s power capabilities. Could it run a bit more silent on lower rpms? It probably could if fed with cooler liquid from the rad, bonus would be less overall pump heat, one of Tom´s complaints. Also the nice looking mesh on top could be enabling an echo chamber effect on noise on colliding frequencies (or not!). A different dampening material as well as/or the apg drive II finned block build would probably help

The hoses should be shorter ? I would say yes by a bit. The longer the travel the less time liquid is being efficiently cooled at the rad. Hose run cooling is totally negligible. I would trade any day one mount option (front?) for a slightly shorter run

Rad seems pretty well built. Copper fins are great if well cooled, if not they could even be worse on a few scenarios. So...

So we come down to the fans. Of course I can understand Tom as he tries to review the kits as they come. But.. as linustechtips shows with a different fan better results are obtained.

To stop the AIO wars (lol) maybe the solution is way easier than expected.

I would say Gabe take notice maybe an improved fan design (or fan option to purchase instead of, call it extreme cooling fan option, whatever) and slightly shorter hose runs for H2O rev 2 would fill in the gaps

note: I am waiting for the next few developments or updates to buy one H2O myself

best rgds

 

[s3ns3]

Its a shame that swiftech use those weak fans.they brag about how silent it runs and thats the point really but only an overclocker would buy these kind of coolers anyway and they do want performance.who cares about silent when we benchmark.anyway ,this is the best performing predone liquid cooler there.i think ppl will buy along with it other fans.it doesn`t have to be noctuas lol,and it does offer other options so after me...best buy

 

[peanutsrevenge]

So, as known (by non fanboi/haters) TLL didn't fudge the review for Corsair

I just got done watching the Linus review and jumped around the FB page and here to see if it was mentioned and OMG, TTL himself has pointed to another reviewers results.

I felt something was wrong with what was tested by OC3D, whether it was a dodgy pump, flaw in the rad or crappy fans, something had to be wrong, surely.

But WTF, they've shipped a product with completely shite fans, how dumb are they?
If you've not got decent fans yourselves, don't ship with fans and let retails make bundles/customers pick their own fans, or pull in another manufacturers fans that actually work and use them.

Poor show Swiftech, these latest findings have devalued you further for me than just the OC3D test where I thought it was a faulty kit.


Must be nice to be TTL with these findings though so the haters can go swivel on further evidence there wasn't foul play

EDIT: I've now gone through and read a little more of the thread and seen Gabe's responses.
Although I don't fully agree with all the information provided (200rpm ~2 degrees), it's good to see his presence here and that Swiftech are hopefully taking on board the concerns, critisisms, suggestions and general feedback by us nerdy bastages. Several + points back to Swiftech!

 

[Ghost_X]

Now I don't understand if the CES presentations were using stock fans or custom fans. Using Gabe's thinking of using everything that comes stock with the coolers the CES results now seem wrong.

Good to know how much difference fans can make though!