IFX-14 Navig's Review

[Ace-a-Rue]

Alright, Part 3, Heatsink Comparison is completed at the main thread.




Here are the summary tables:

Thermalright IFX-14 vs Ultra 120X, Single Fan Configurations



Notes: all testing for the IFX-14 was done with the fan in the middle position (between sink towers)
and for the Ultra120X in the Pull configuration (Pulling air thru the tower sink).
Flames indicate where the cpu failed due to overheating. 75c probably doesn’t represent an accurate temp, as the cpu automatically terminated its load.


Well, with a low speed fan, the Ultra120X actually failed. Its not so surprising as the heatsink tower has fins that are very dense, as well as large in surface area. The IFX-14 appears to have the same density, but breaks the surface area into 2 towers.

However, once you get some fans with some oomph, the 2 heatsinks performed virtually identically. Maybe the slightest edge to the Ultra120X with 38mm thick fans.







Notes: Best Mod mode for the Ultra120X was Push-Pull Panaflo M1's. Best Mod mode for the IFX-14 was Delta25 - Panaflo M1 - Panaflo M1 + 80mm fan on the backside cooler.

With 2 fans, the Ultra120X gains a small edge.

However, in the best mod mode, you can strap 4 fans to the IFX-14! The 2 best mod configurations are pretty equal in sound, but the IFX-14 gains a 1.3c advantage. The ability to use more fans increases its performance, with minimal addition to noise.


IFX-14 vs Ultra120X comments

Well, I think its much more complicated than saying one heatsink is better than another. They always performed within 1-2c of each other. As of the writing of this article, the IFX-14 is some $30 more than the Ultra 120X. If you are looking for that extra money to translate directly into higher performance, you will be disappointed.

What DO you get for that extra money.
1) A much better mounting mechanism.
2) Significantly better performance if you are looking at using only ONE low speed fan, and generally better performance for low noise solutions (for example 3 yate loons is quiet yet performs quite well).
3) Native support for multiple fan configurations (Recall the Ultra 120 requires modding to support a second fan).

There may be other differences, which I haven’t specifically pointed out. For example, with my IP-35, the IFX-14 consistently beat out the Ultra120X in PWM temps.

On the downside (besides the price), the IFX-14 is enormous, which may interfere with large gpu coolers, nb coolers, or active RAM cooling.

interesting results compared to Vipor's results...i would say significant difference...he showed the Ultra-X to be ahead in almost every category...so....i have to interpolate your results with mixed feelings as to whether the IFX-14 is worth the expenditure to get equal or slightly less performance than the Ultra-X if you use the right fan(s) with IFX.

another interesting statement; you said, you got lower PWM temps using the IFX...
----------------------------------------------
my questions to you:

"how did you mount the IFX and Ultra-X"?

did you use the secondary cooler which is mounted beneath the mobo?

were the heatsinks facing to the rear of the case (i presume all testing was actually done on a bench, outside of a case)?...OR...were they mounted in the direction of south-to-north direction as the board would stand in the case, in a vertical position?
----------------------------------------------
if you mounted the secondary IFX cooler that would explain lower temps for the PWM...the sensor is adjacent to the cpu socket on the side that would face the PSU when installed in the case....you can put a 40 or 60 mm fan right there pointing at the cpu socket and influence the PWM temperature reading.

here is my final thought...the IFX is a massive cooler...but..."NO" cooler will operate at full stress load (especially the quad cores unless they are the newer "G0" steppings) inside of the case with the side door closed and you are running a high overclock.

i have a case that has two rear 120 mm fans mounted to exhaust...they are running at 2.5 RPM (silverstone) in the exhaust position (besides the 120mm in the PSU)...they cannot keep up with extraction of the super-heated air coming off from my Ultra-X with the side door closed when running Prime 95...as soon as i take the side door off...no problem....SO...it will not matter how many fans you mount on the IFX...heat is heat!...trapped inside, it will be a hinderance for a full stress load...it is why water works so much nicer than air heat exchangers!

EDIT:..forgot to add that you can mount two 120 MM fans on the Ultra-X by just purchasing an extra set of clips from thermalright...they can be hooked into the fins just one notch below the normal hook up for one fan.

[happ]

i..forgot to add that you can mount two 120 MM fans on the Ultra-X by just purchasing an extra set of clips from thermalright...they can be hooked into the fins just one notch below the normal hook up for one fan.

Have you a link where such a clips can be purchased?

[lad]

Hi, Navig.

You have gone to extraordinary lengths to test and publish your findings, and we all appreciate it tremendously.

I think the IFX is cooling less well than expected due to hot air from the first 'bank' flowing through the second.

Since you have such a great deal of reliable data, would you be willing to try one more fan mount on the IFX?

That would be to mount two fans, each one pulling cold air from the open center fan position. If I understand correctly, your are running caseless, so you won't be faced with hot case air feeding into the IFX.

I bet you get another 2-3 degrees with this layout!

Thanks very much for all your dedication and work!!!

[Ace-a-Rue]

Have you a link where such a clips can be purchased?

i just looked again at thermalright's website and they show the 120 clips for the "XP" series...you might want to email them and maybe they will sell you the Ultra 120 clips...i did that a while back about the K8 back plate was listed but no price affixed....they sent it to me without even asking for payment...i mailed them a check anyway to cover the cost and postage....they are a great company.

here is what they show for 120 mm clips:

[Navig]

Ace-the lower pwm temps are obviously explained. If you hit up my oiriginal thread, there is much more information--I only summarize the results here. All testing was done with the IFX-14 mounted so that exhaust would be over the i/o port, airflow parallel to the video card. That's the only way to support 38mm thick fans on the intake side of the heatsink. If you put fans in the intake position, or in the middle position, the fans hang below the heatsink and generate a flow of air under the heatsink, that runs over the pwm coolers. On the Ultra120X, this flow exists, but is minimize because the fans naturally sit much higher (because of the clips and the mounting mechanism), causing there to be airflow above the heatsink instead of below.

I don't own any standard cases, so I can't really test that. These parts are destined for this case here, so we'll see how temps turn out. Testing this heatsink was a bit of time coming, but a) I work a full time job b) my modding earns me money whereas this sort of thing is done purely as a service.


Lad-I'll try it, but I doubt it. The air temp in one pass thru a heatsink is only minimally changed. If I reach over and feel my exhaust, it is only a tiny bit higher than ambient. A lot of people feel their case exhaust and it is very warm, but remember, I'm working with an open bench. Most case exhausts are warm because they are recirculating air in the case because of inadequate exhust. See Ace's post above. On the other hand, air pressure is a huge factor, and aiming 2 fans at each other cause a big effect in reduced airflow. I suppose I could bend the heatsink towers at an angle, but then the fans would no longer fit.



Thermalright's a great company, I think my first Thermalright was an sp-94. Like the Ultra 120X and the IFX-14, I picked up an XP-120 the day it was released.


Comparing aircooling to watercooling is useless. If you look at my case mod link above, its obviously set up for watercooling, but I got stalled out trying to put together watercooling parts--have to modify my pump, one block got lost in the mail, used radiator's rusted. I still have plans on making this case watercooled--thinking about running clear pvc hard lines!

[Ace-a-Rue]

MAN...what a case!...that thing looks like my stereo console!

[happ]

i just looked again at thermalright's website and they show the 120 clips for the "XP" series...you might want to email them and maybe they will sell you the Ultra 120 clips...i did that a while back about the K8 back plate was listed but no price affixed....they sent it to me without even asking for payment...i mailed them a check anyway to cover the cost and postage....they are a great company.

here is what they show for 120 mm clips:

Yes, I have seen it, but I am not sure will it fit to TRUE 120. And I don't think they are going to send me right clips all over to Croatia.

Thank you anyway.

[Ace-a-Rue]

Yes, I have seen it, but I am not sure will it fit to TRUE 120. And I don't think they are going to send me right clips all over to Croatia.

Thank you anyway.

yeah, i think thermalright does not ship internationally...someone from cairo was trying to get a K8 metal back plate from them and he had indicated that they won't ship it to them.

[Navig]

Alright, Lad this is for you:

Panaflo M1 x2, mounted like so:

Both fans blowing in the same direction.
Idle:
Uguru: cpu 41 / sys 30 / pwm 44
Coretemp: 42.0

Load
Uguru: cpu 57 / sys 30 / pwm 56
Coretemp: 64.1


Now the same configuration, but both fans pushing air inwards:
Idle:
Uguru: cpu 41 / sys 30 / pwm 43
Coretemp: 43.5

Load
Uguru: cpu 58 / sys 30 / pwm 53
Coretemp: 66.0


navig

[Rolle2k]

Alright, Lad this is for you:

Panaflo M1 x2, mounted like so:

Both fans blowing in the same direction.
Idle:
Uguru: cpu 41 / sys 30 / pwm 44
Coretemp: 42.0

Load
Uguru: cpu 57 / sys 30 / pwm 56
Coretemp: 64.1


Now the same configuration, but both fans pushing air inwards:
Idle:
Uguru: cpu 41 / sys 30 / pwm 43
Coretemp: 43.5

Load
Uguru: cpu 58 / sys 30 / pwm 53
Coretemp: 66.0


navig

Navig: Its exactly the same fans that you had at the Panflo Multiple fan test up above? if so, then the temprature 64.1c is much better at load than the u120x right? Cus at the other 2xpanflo test you did above the ifx-14 had 67.9c and the u120x had 67c.
Also on all the other multiple fan tests (except this last one), how did you mount the fans? side blow, and middle position?

thanks for testing all for us

[Ace-a-Rue]

Have you a link where such a clips can be purchased?

i do now...just got it and they ship internationally!

http://www.xtremesystems.org/forums/...d.php?t=154104

[Navig]

Everything is well explained in the original thread, in regards to setups etc. The reason the temps are so much better is because the ambient temps today are >10 F lower than when I did the rest of the testing. Underscores some critical elements--when talking about the top tier of heatsinks, were are talking differences of 3-5c under controlled circumstances. Other factors like ambient temps, case air flow can easily overwhelm these differences.

navig


ps. I live in Boston--no AC. Day before yesterday during the main testing it was a sweltering 90f (32c). I typically do my testing all thru the afternoon, or all thru the middle of the night, to have consistent ambient temps. This afternoon, its only 80f (27c).

[lad]

Lad-I'll try it, but I doubt it. The air temp in one pass thru a heatsink is only minimally changed. If I reach over and feel my exhaust, it is only a tiny bit higher than ambient. A lot of people feel their case exhaust and it is very warm, but remember, I'm working with an open bench. Most case exhausts are warm because they are recirculating air in the case because of inadequate exhust. See Ace's post above. On the other hand, air pressure is a huge factor, and aiming 2 fans at each other cause a big effect in reduced airflow. I suppose I could bend the heatsink towers at an angle, but then the fans would no longer fit.

No need to bend your HSF any more, Navig!

You raise a good point about case temps. I just measured mine...
Idle core: 37c, 35c
Ambient: 77f
At intake to lead IFX fan: 78f
At exhaust from IFX fan: 83

Orthos core: 65c, 65c
Ambient: 77f
IFX intake: 78f
IFX out: 90c

So at ~800 rpm / 7 volts, I gained 12f thru both towers.

I'm tempted to throw my 'ventilated' side panel on, an do two exhausts on the IFX, and put one Panaflo exhaust in my 5.25" bay... But I know I wouldn't want to leave it that way, as the fan noise from my video card and PS and HS would annoy me too much.

I *did* try a cardboard divider in the center slot, to guide hot air from the front half of the IFX into my PS, but that impeded airflow too much to work.

[jalyst]

Hey people, sorry to be a lil off topic but I just gotta know.....

Has anyone fitted this heat-sink to a Gigabyte P35-DQ6 (rev 1)?
This motherboard uses an overly elaborate NB and Mosfet heat-sink.

http://www.tweaktown.com/reviews/110...ere/index.html

I'm concerned that I may not be able to use it with the IFX-14! :-(
There's a newer revision with a more streamline heat-sink but I already bought my P35-DQ6 6wks ago*

Cheers,
jed

*haven't unpacked it yet, but I just know I'm going to have probs if I try to bring it back

[ineedaname]

Woooooooot nice review.

Couldn't wait to see the results on those delta's

[Navig]

Sorry, Jalyst, I just don't know. If I had to guess, tho, it probably would. The heatsink fins sit pretty hi.

I will note here, that I most definitely got better temperatures with pressure mounting, when the motherboard is in a vertical position.


This is one of my current systems, with the same e6600 as tested:





As you can see I have easy access to the motherboard. Well, I noticed temps were okay, but maybe a shade higher than they ought. Having never tested the IFX-14 is anything but a test bench, I thought maybe there is something to this pressure testing. So I opened the doors and simply pushed on the heatsink towers, towards the top. Pow! Instant drop 2-3c off load (smp folding). Let go, it jumps right back up.


So I brought it all out, and added 2 levels of doublestick foam tape between the hold-down plate and the heatsink base. And now temps were 2-3c lower, with no difference if I press.


navig

[Spawne32]

Is that a fish tank??!?!

[Navig]

Whoops, should have linked it.

[Spawne32]

Whoops, should have linked it.

by far the finest custom case i have ever seen

[jalyst]

Thanks Navig, think I'll bring my mobo into the retailer and check 1st!

So, from a pure 'cooling performance' perspective (noise it not an issue)...
Which heatsink do you honestly feel is best?

I've read through the entire thread (along with vapors) & you both seem to have slightly conflicting opinions.
http://www.xtremesystems.org/forums/...=152773&page=7

Vapor feels for raw cooling performance (noise not a factor) 120x is better.
Unless you're getting high into OC'ing territory & only then may the IFX-14 start to get a slight edge. Whereas you feel that because of the 4-fan mod, the IFX-14 is a better high TDP cooler.

So I'm no wiser and even more confused?!!? :-)

Cheers,
jed

[jalyst]

*bump*

Navig, did you see my last post?

cheers,
jed

[PhilDoc]

Great and very informative review. I assume the Delta 38 is a VHE. Interesting that the single VHE out cools 3 lower cfm fans. Of course noise is a huge factor. Three VHEs on a fan controller looks the most interesting to me. Thanks again, this review really helps with deciding what to buy.


Btw, as has been already said, just an outstanding case.

[Navig]

jalyst, I wouldn't lose sleep over it. I have both and I don't. I know vapor has both and he doesn't. Both are excellent heatsinks performing very very close to each other. Probably variations in thermal grease would yield bigger differences. Personally, tho, I think that if you strap 4 high flow fans to the IFX-14, it'll probably outperform.

[Vapor]

I'm using my IFX-14 on my E6700 and my Ultra-120 eX on my QX6700....with a modded mount on both, they're very competitive and 1-2C in either direction is no biggie.

I'm still kind of in awe that the IFX-14 doesn't walk away with the lead, but on closer inspection, it begins to make sense

[jalyst]

Thanks guys.

Sorry for being so anal but I just don't have the time/money these days to prove definitively which is better for my intended application.

The IFX-14 seems to be avail for the same price in most places down here in Oz. So, as long as it cools better with 4 high CFM fans than the coolest mod mode of the 120x, I'm in like Flynn!

Sorry to push it but Vapor are you in agreement with Navigs last sentence? Your previous response was very "diplomatic".

cheers,
jed