Very nice, can't have too many compact coolers, especially non-slot-hogging ones. I just wish this had been available before I recommended Scythe to a friend, who then proceeded to snap a screw in the rather flimsy mounting kit. Noctua's always seems to be pretty brute-proof.

This would be great for really small mini-ITX cases. Also great for smallish mini-ITX gaming rigs.
Too bad there aren't any decent (or at all?) low profile video card coolers

What would be the point of installing such a low level cooler if you'll have a videocard towering over it anyway?

Because I was talking about low-profile video cards you silly

http://img19.imageshack.us/img19/1082/sam0297s.jpg

or full size video cards on a angled riser, in a low profile case!

There is no clear definition of "low profile" but at 37 mm, it should fit in most consumer SFF cases that described as such.

But there are clear height constraints specified in the mini-itx form factor definition. http://www.formfactors.org/developer%5Cspecs%5Cmini_itx_spec_V1_1.pdf - which given the position of the CPU socket in most Sandy/Ivy Bridge mini-itx boards is 38mm (it would be 58mm if the sockets were placed more centrally). Hence this heatsink looks to be compliant with mini-itx standards - whereas the Scythe Kozuti at 40mm isn't.

And yes I have a mini-ITX case that exactly meets the standards with a 38mm clearance between the motherboard and the drive cage, so hopefully this may fit.

FYI, I am currently using a cheap and cheerful Asaka low profile cooler (there seem to be several similar models) which happily spins at an inaudible 700rpm at low CPU usage. I also tried the Gelid Slim Silence, but found that this did not live up its name with a minimum fan speed of 1400rpm.

Welcome to SPCR -- a good first post! Thanks for digging up the mini-ITX FF spec. I hadn't realized Intel had actually done up one of these, as it was a VIA adoption of a mobo FF originally created by Shuttle.

Anyway, I think you misread the illustrations. The 38mm height restriction you refer to is for the back panel I/O area, specifically a strip 27mm from that back edge. None of the Mini-ITX boards we know about put the CPU socket that close to the I/O panel, so the 58mm max height applies. If there was a spec for Thin-ITX, 38mm might be appropriate... but AFAIK, there isn't one, so for now, we live with the vagaries of "low profile".
Surely that min speed is a fan controller limitation, not the fan itself.

Sorry Mike, I think you're the one misreading the spec. You're referring to area D, which in fact has a height restriction of 39mm, a whole 1mm higher. I'm referring to area C a 30mm strip at the "front" of the board away from the back panel, which has a height restriction of 38mm. The socket area of many Intel M-ITX boards fall into area C. I'd agree none fall into area D.

I'd also agree that I haven't seen a low profile M-ITX case that claims to be standards compliant. I've seen a few with a 38mm clearance between the board and the drive cage, but most intrude into the 58mm zone.

(BTW There is a standard for Thin iTX - its hidden in the latest version of the M-ITX spec. http://www.formfactors.org/developer%5Cspecs%5CMini_ITX_Spec_V2_0.pdf. But thats a different can of worms).



It was very much a function of the PWM Circuitry within the fan Gelid used - they specified a minimum of 1400rpm and no lower would it go, even though it reached this speed quite some way above the minimum PWM values available on both Intel and Asus motherboards. e.g with Asus Fan Expert, it reached 1400 rpm at a value of 50, setting the PWM value down to 20 had no effect whatsoever.

Another Noctua fan that doesn't live up to it's specifications :X

I've have been waiting for SPCR to perform a review of this cooler since I saw a little while ago. I wanted to know if it was worthwhile replacing my stock 1155 cooler (I suspect so) with the NH-L9i and was hoping to see a direct performance comparison.

This information may already be on the site somewhere but I couldn't find it so I was wondering if someone could show me the performance of the stock cooler on the same testbed.

Thanks also for all the high quality reviews. SPCR has been my go to site for all my cooling information.

OK.

Just checked on a couple of 1155 m-ITX boards, and they are both very close to the 30mm wide area C, but not less than 30mm. I would be surprised if the boards actually put the CPU socket closer than that to the edge, there'd be no room for the RAM slots which normally go there. But even if they do, since neither coolers nor cases adhere strictly to the m-ITX spec either, it's all kind of hit and miss.

It's not just the low profile cases which leave little room for the CPU heatsink, but also the ones that position the PSU above that area. If it's an ATX PSU, it can cover the whole length of the board.

But this business of the case not being staying w/in spec also holds true for ATX and mATX case design as well. The actual ATX case spec itself does not reflect all the innovations that have been implemented in cases that hold ATX boards, and these innovations have actually kept ATX viable for all the power/thermal increases that the original design was never meant for.

For sure.
That does seem silly. We would have missed that because its speed was voltage controlled for our review; this was before Fan Xpert 2 on the Asus P8Z77-V Pro was adopted for PWM fan testing.

My experience is that you can, to some degree, adjust the working speed range of a PWM fan by reducing the voltage it's fed. I first did this some way back an a Zalman 9500AT cooler that had 1200rpm min 92mm fan that wasn't quiet enough. I use a Fanmate wired in to just the +ve and -ve wires leaving the speed signal and PWM control wires untouched. You can then finely adjust the voltage supplied to the fan to "move" it's working range.
For example the TY-140 fans I use normally operate in 700 ~ 1300 rpm range but with some voltage tweaking they now run in the 500 ~ 1000rpm range on 30 to 100% PWM.

Scythe now sell a fan controller that does exactly this.

Regards, Seb

Thanks for the review!

Can the heatsink be rotated or mounted in any orientation? I noticed in the photos you have it installed so the fin openings are directed toward the IO ports and memory. I'm hoping it will also mount rotated 90 degrees from what your review setup shows?

Not positive. The ends of the heatpipes have nubs that protrude on one side, so this is probably motherboard-dependent. Why is this important? The amount of peripheral around the cooler hardly varies at all, so there's not significant advantage for cooling any other board components.

Ya, that was my primary concern, whether there would be more component cooling out the open sides for nearby components. I think your right, not really an issue. I've gone ahead and purchased this cooler to give it a try. Thanks!

While I haven't yet seen one in person, the one thing I don't like about Noctua is the brown color scheme of there fans. Wish they'd change to black.

If the heatsink's fins are super tight I wonder how a fan with high static pressure would do. Dare I say 92mm Gentle Typhoon?

@ SPCR: I don't understand your conclusion about acoustics. The table with the acoustic results clearly shows the Kozuti is much more quiet (check the Kozuti review here at the bottom). At 8V it makes same noise as Noctua NH-L9i at 6V. So especially when idle and low/medium load, the Kozuti should be more quiet right?
Only on (heavy) load the sound of the Noctua is better then the sound of the Kozuti. Right? But in the conclusion, you state the Noctua is more silent


I am looking for a fan for a HTPC based on AMD A10-5700 (65W) wich is only 25W TDP on idle and can be even lower:
http://hexus.net/tech/reviews/cpu/47257 ... 00/?page=8

But it will have to fit in this case: http://www.lc-power.de/index.php?id=122&L=1
Wich is very low profile.

I received my NH-L9i yesterday.

It can be rotated four ways (as long as the wires reach the fan header). The only problem I ran into was a slight manufacturing inconsistency. One of two heatpipe nubs protruded slightly past the keepout area, and interfered with the 4 pin ATX 12v connector. Not really a big deal, as the position I went with wasn't affected by this (open ends pointing left to right, and nubs toward the io ports).

Coming from a Big Shuriken 2, the NH-L9i has a noticeably higher base idle speed (PWM controlled on the same mb). Not sure why this is? It's around 1025 rpm. The Big Shuriken 2 will idle for me under 1000 rpm. I'm running a Pentium G620. So obviously the NH-L9i seemed slighly louder than the Big Shuriken 2, but then it is spinning faster, so it's not a fair comparison. Despite this, it is still very quiet. Pretty much silent for me at 1 meter away with an open case (and about 1025 rpm).

You know it's quiet when your monitor is the loudest component!

Thank you! Very very good review
Read the review and thinked "what if I would slap a 92mm GT on this thing" and came to read this thread and noticed herby asked the same thing.
I happen to have a 92mm GT (1700rpm) sitting useless in my stuff box..
Should be quite good this kind of cooler, the static pressure I mean?

The reason I'm interested in this cooler is because I'm upgrading my HTPC to S1155 era and moving from mATX/ATX MB/case to mITX MB/case.
I already have an Asrock B75M-ITX, Intel G2120 and a 4gb stick of DDR3/1600/CL9 coming in. Next step would be to buy a new case. After much reading and browsing and reading I have decided on Lian Li's Q11. With that I can keep most of my stuff. Only CPU cooler needs to change. (HD5670+Accelero S1r2 will be interesting to fit but I have a dremel if I need to make the cooler smaller lol)

jamese, what kind of load temperature are getting with your NH-L9i + G620 combo?

E: LOL at my signature. Been a WHILE since I have "boinced" last time. Also is there a way I can change my username?? I'm lagittaja on many many forums and I have no freaking idea why I have registered here using this nick