Advertisement

The AK824cu is the latest offering from Akasa, the "manafacturer of award winning Silver Mountain cpu cooler" as they proclaim on the packaging. [ Silver Mountain | Silver Mountain 2 ]. This is a more standard design than the rather unique Silver Mountain range - the tall thin pins of the SM coolers have been replaced by a standard extruded fin design. The material has also changed - the rather extravagent silver plated copper of the SM has been ousted in favour of good old traditional aluminium. With all the hoo-haa associated with using copper and other fancy materials it's nice to see a normal aluminium heatsink. The fact remains that aluminium is one of the most efficient materials at dissipating heat.

This sink isn't entirely aluminium mind. At the base of the sink is a solid block of copper, that celebrated heat conductor. I like the idea of copper bases and aluminium fins... it seems to make sense in theory - a great conductor at the base to sap the heat off of the cpu die, and then an efficient radiator made from a great dissipating material. My knowledge of these things is sadly quite lacking and based on word of mouth rather than education, so if anyone wants to tell me how wrong I am please feel free to do so.

Those clever Akasa bods have repeated their successfull decision from the Silver Mountain 2 to include a 3-lug-thing clip. What I mean by that is the clip attaches to all 3 of the sticky-outy bits on either side of the cpu socket. This means a much more secure fit, more even weight and pressure distribution on the cpu die, and less danger of snapping off the plastic clips on the cpu socket. (I wish I had a proper word to describe those bits with). It's shocking how many coolers still use the crappy 1-notch clip design. I don't understand why everyone hasn't adopted these decent clips. I hope Akasa will continue with them (although with new chip generations this clamping method will become obsolete).
The only odd thing about this clip is that the screwdriver notch (for seating the clip) is on the inside of the clip. Looking at the picture will just about reveal what I mean. This does in fact make it a bit easier to bend the clip out and over the socket lugs. Which also reminds me to mention that getting this thing on is not for the faint hearted. You really need to put some force in, which is slightly worrying when dealing with you pride and joy expensive hardware. Once it's on, though, you get the feeling it's not going anywhere.

The overall design of this heatsink seemed very familiar to me. In fact, it's almost identical to the Spire WhisperRock (pictured right) which I reviewed some time ago along with 2 of its brothers. Very, very similar... I wonder if there was a bit of inspiration involved in the design of this cooler. Anyway, they say imitation is the sincerest form of flattery (or words to that effect) and since the WhisperRock performed pretty well back then the design similarities can only be a good thing. The Akasa offering also brings a lump of copper to the party, which will hopefully improve on the proven design. Spire have since released a third revision of their cooler, an example of which I might try and get hold of.

The back of the (rather attractive) packaging states dimensions of 80x80x36 for the heatsink and 80x80x25 for the fan. The base of the heatsink is rather smaller than the top, the heatsink tapers away quite significantly. This is good as it allows for the mistakes made by daft motherboard manufacturers who see fit to tuck capacitors and such in next to the cpu socket. There were no installation issues on my superb AOpen AK77-400MAX, but then the area around the socket has been kept free and clear by those clever AOpen engineers. Did I mention I really like AOpen?

The packaging spiel also gives details of the fan as 2500RPM rotational speed, 32.37 CFM (cubic feet per minute - what a daft measurement) airflow and 27.85 dB(A). For those who don't really get the whole decibel system (and to be frank, I find it pretty damned confusing) here's a simple few examples of noise levels I found on my travels around the net:

0 dB is defined as the THRESHOLD OF HEARING, and it is with reference to this internationally agreed upon quantity that decibel measurements are made. In some situations, such as tape recording, a given intensity level is assigned 0 dB, and other levels are measured in negative decibels in comparison to it.

I'm not completely sure about these figures and whether they truly relate... if you know otherwise, please drop a note in the forums about it.
This table gives us a noise level in between rustling leaves (a bit of an odd one) and a quiet whisper. The quiet whisper I am reasonably happy with, this fan is pretty quiet.

Anyway, enough of the signals theory lecture. On with the matter at hand. As with most of the recent Akasa heatsinks, this came with their "ShinEtsu" thermal interface material. Most people will agree that generally TIMs are crap but back when I was sent the Silver Mountian 2 I was told that Akasa were confident that their ShinEtsu pads would perform just as well or better than more expensive thermal compounds such as the ubiquitous Arctic Silver range [ Compound roundup | Nanotherm compound ]. Well the ShinEtsu didn't fare too well during my Silver Mountain 2 review and unfortunately things didn't go much better with the AK824, although the fault lies with me this time.

After installing the cooler with the TIM attached my computer refused to POST, so I removed it and used Arctic Silver 3. However it seems my computer wasn't booting due to some strange Zip drive problem... but of course after scraping the thing off there was no going back to test it. The pictures show the state of the heatsink base and cpu die after installation - it seems the TIM got quite firmly in between the die and heatsink base since there is no copper visible and the die has some residue on it. Since I didn't run the thing for long, the ShinEtsu probably didn't have chance to change state (i.e. melt) and get into its desired condition. Sorry ShinEtsu, maybe I'll give you a fairer trial next time.

So finally I got the AK824 into action using it's new dob of AS3. The cpu in use is an Athlon XP 2400+ which I think represents a fairly common cpu and thus hopefully a good test bed for real world results. Testing conditions are, to put it mildly, harsh. As you may know us Brits are going through somewhat of a monster heatwave, as are the rest of Europe it seems. Global warming? Nah, that's just greenpeace propaganda... it'll never happen...

Temperatures are recorded by my motherboards onboard sensors via Motherboard Monitor 5, undoubtedly the greatest piece of software of its kind. The rather attractive green LED fan atop the AK824 will be running at full speed since it's so quiet at 12V I see little point in undervolting it. Even the most noise concious among you shouldn't see a need in quieting this fan down. The test system is enclosed within a Lian Li PC60USB case, with 2 front 80mm fans and 1 rear 80mm exhaust (an Akasa LED fan as it happens). The Enermax psu has a 92mm and 80mm fan within, but seems to generate a fair bit of heat itself so doesn't act as a wonderfully efficient heat exhauster.

So now you understand the conditions, how did the thing perform I hear you ask. Well, pretty impressively in my opinion. During the day my room was about 30 degrees C. With an ambient temperature like that, any cooling system is going to be hard pressed to perform well. The cpu remained at about 50-52 degrees all day, and that's at full load (the UD client running and occasional bouts of Max Payne to push things a little more). Late last week conditions were nowhere near as harsh... with an ambient temp of 23.5 degrees C the cpu maxed out at 42 degrees under load and idled at around 39. All this from just a simple copper/aluminium heatsink and a low speed 80mm fan. (The quoted speed of 2500 is a little low, MBM reports the fan as running at between 2600 and 2800 RPM... it seems to vary which I find rather odd.)

Just to confirm things to the sceptics, a screenshot from MBM:

I find these results frankly quite surprising and very pleasing. Again, room temps were around 30 degrees. To keep just 20 degrees above ambient isn't bad at all. I'm also testing a ThermalTake Volcano 11+ at the moment and whilst it managed to hold similar temps during last week (when it was colder), it does so with a higher cost and much (much!) more noise. And it doesn't even have a green fan.

It has just occured to me that I haven't included any pretty pictures of the green LEDs in action. Well, to put that right here's a couple of shots of the fan running amidst my red and blue colour scheme. I think it looks pretty good, and impressively it manages to shine through the bright red and blue light sources quite effectively.
Purdy huh?

Right, I think I've exhausted things to say about the AK824cu now. So it must be time to conclude.
All in all the AK824cu is a very acomplished cooler. In a world of crazier and crazier cooling solutions being touted by the various companies, it's nice to see things can still be done the old fashioned way. And to spice up the "normal" look, the green LED fan spices things up nicely. Of course those of you without case windows won't get much out of the LED fan, so for you people there's a version without the snazzy coloured lights. How nice of Akasa.

The noise generated by the fan atop the AK824cu is refreshingly low. A lot of new coolers, the aforementioned Volcano 11 for example, include stupidly high powered fans. Whilst these things might be able to push a bit more air through the inefficiently designed heatsinks (ooh, get me!) the noise generated by these horrible squealing fans is unbearable, to me at least. The low pitched hum of a slow 80mm fan is far more bearable than the 60mm monsters that some heatsinks use (and in fact the 80mm monster than the Volcano uses). Akasa made a good choice with the fan on the AK824cu and clearly designed the heatsink sufficiently well that low airflow saps the heat away efficiently. On a side note, it seems wider spaced fins are better suited to low airflow situations, while the tightly packed copper fin coolers like the Volcano and other coolers seem to prefer high powered fans.

I highly recommend this heatsink, especially due to the one detail I have kept from you thus far - the price. Heatsinks command quite silly fees these days - the much vaunted Thermalright SLK800 costs a mastercard melting £35-40. The 900 model is another £5 on top of that. The Akasa AK824cu on the other hand (that's the one I've been babbling on about) costs a much more wallet friendly £15 with a spangly LED fan or £13 without it (approximate prices). With a price like that, its an excellent buy for system builders and performance users alike. Sure it might not do quite as well as the more ridiculously priced coolers but I think for the majority of us this represents an excellent option.

Well done Akasa.

Good performance
Great price
Built solidly
3 point clip
ShinEtsu should remove need to buy expensive paste
Low noise
Funky coloured fan (green or blue) - all important
Tight, and slightly scary, clip
Possibly not adequate performance for the true hardcore

Related:
Akasa UK
Akasa Taiwan