ASUS Maximus VI Gene uATX motherboard, Intel i7 4470K CPU, & AMD FirePro W5000 graphics. So far I've found the Silverstone NT01-PRO, Scythe Big Shuriken, & Gelid SlimHero as likely best options. No real over clocking intentions (K CPU was on sale) but system will be used to potential, not a HTPC.

NT01-PRO is appealing because it is designed for this case. I would try case fan only but add fans if necessary if they fit. Considering custom bracket for 120 mm downdraft fan mounted below optical drive only if necessary. Other options are downdraft too but is that really best in this case? My gut feel is Silverstone solution just makes more sense (heat rises, case is small, air flow considerations, etc.).

Thoughts?

Unless you undervolt/underclock the CPU, the NT01 Pro can't safely handle a 4770K passively (even at stock).
It does need at least an 80mm fan, but it would be better if the fans are two, and even so it would struggle to outperform smaller designs such as the NT06 Pro, the Noctua NH-L12 or the Scythe Big Shuriken 2. The Prolimatech Samuel-17 with a Prolimatech UltraSleek Vortex, and the Thermalright AXP-200 might be two valuable alternatives also, even if the most bang for the buck it's still the Big Shuriken 2, thanks to its remarkably good sounding stock fan.

Eventually, IMO the NT01 Pro is an interesting design, but it wouldn't seem to worth that much.

I should add that I fully intend to keep (2) 80 mm case fans, stock Silverstone for now but will replace with something better (about the same or better CFM, but hopefully lower RPM & more quiet). I'm still undecided on exactly which fans as it may depend somewhat on how I wire them & whatever other cooling goes inside.

So technically the NT01-PRO would not be operating passively, right? Or are Silverstones's thermal specs always assuming at least (2) case fans nearby? My assumption is they truly mean passive for the 65 watt spec - no fans at all.

Also, these down-facing cooler/fan combos seem counter-intuitive from the start don't they? Heat rises after all. They seem a bit more counter-intuitive to me for such small cases & more so in this case because the case fans are very close & blowing out. Wouldn't this create a pocket of low pressure basically above the CPU? The case is not ventilated well, especially up high. Wouldn't it work better in this case if these CPU fans pushed upwards so cool air would draw from below (where there is better case ventilation or I could add holes easily) & so the case fans could more directly expel the warm air?

I'll check out the other option too, thanks!

Well, no, I'm pretty sure the correct assumption is that they mean "no fans on the cpu fins", but case fans running nearby.
Moreover, the GD02 looks like less versed than its GD04/05/06/07/08/09 siblings, in order to run the NT01-Pro "fanless".



Well, no I mean: it's just thermodynamics.
Just to give you an image: when you want to sip a cup of tea, or a mug of coffee, and the liquid is a bit too hot, when you do not want to burn your lips, what do you do? Do you blow on the surface of the liquid, or do you suck the hot fumes? Even if "Heat rises after all", you blow (and don't suck): it's the most effective way to cool down your favourite liquid.

That's probably true since the NT01-PRO appears to have been designed specifically for these Silverstone cases but I still have my doubts. I suppose the only way to know for sure would be to ask Silverstone.


How so?


I realize that but I suspect in a small case the actual flow of air into, thru, & out of the case plays a significant role. In your analogy, the stream of cooling air from your mouth is very directed blowing vs. sucking open air above the cup. Enclose that cup with a close fitting lid & straw & the dynamics are completely different!

In the case of the fan & radiator I bet it makes little difference (to the CPU) which direction the fan blows except that downward flow will spill over to possibly aid in cooling other components nearby. However, that downward flow will also disrupt air flow into, thru, & out of the case. The net result could be a wash. I suspect the only way to know for sure is to test.

Does anyone here have direct experience with this case?

Because of the larger case fans and the closer to them CPU placement, which you will have on those siblings.




Well, no, AFAIK it's mostly a matter of energy: given the same ambient, you'll supply more blowing than sucking, easing the heat transfer from fins to air.




It's the scientific method: however, as this morning it comes in handy for me to pick a shot, look at this picture:



It's an AMD Phenom II X4 960T at stock, inside an Antec Micro Fusion Remote 350, cooled by a Scythe Big Shuriken 2, while running a Prim95 Small FFT session.

The left side of SpeedFan graph shows the cooling prowess with the fan blowing from top to down (the canonical way): it tops around 45°C. The right portion of the same graph shows what happens when I flip upside-down the heatsink fan, so blowing from down to top (I mean, sucking the hot air from the heatsink fins): it tops around 55°C, about 10°C worse.

Sound-wise it worths to mention that in the specific enclosure (so a slim HTPC-style desktop case) this flipped setup is fairly noisier than the straight top-down blowing one, as the CPU fan acts like an intake case fan where the heatsink is a fairly restrictive, high impendance grille, thus generating a lot of airborne noise: moreover, the top case panel, pushed by the airflow coming from the CPU fan nearby, also resonates.

Well, good luck with that Grandia, and have a nice Labour Day!

It doesn't surprise me that flipping the fan can make a difference. The magnitude of the temperature difference is surprising though. Acoustics are another issue & one that I do have some experience with professionally. I suspect it is often easier blowing than sucking simply because the components themselves form a natural guide for the airflow. Do you have a link or any more details on this experiment? I'll check out the case, but other details are equally important.

The Grandia GD02 is a challenge for sure but I really like the case & I've had it a long time. Others like it are just so ugly! I'm not after perfect cooling, just reliable & as quiet as possible without sacrificing computing performance. Probably a challenge in this case.

It looks like some experimentation is in order. A Big Shuriken 2 is hard to find & a bit expensive. The Gelid SlimHero is inexpensive & readily available. The NT01-PRO is also expensive, especially if I add two more fans (which seems likely necessary). I may have to do this in the name of science though.

Thank you for sharing your thoughts!

I think there's a misunderstanding: in order to show you some actual data after my words, I made that test right in the morning, and then attached the relevant screenshot to my post (look at traybar clock: I took the shot some minutes before posting).

Anyway, that result was rather obvious, and if you think about, in any kind of CPU air cooler (tower, top-down, whatever) the fan blows through the fins "pour cause" (to be fair, you can have a fan aspiring from the fins just in a push-pull setup, where the fans are two).




If you want to stick with the GD02, there's no problem, it has its own merits.
If you want to stick with a rather useless Core i7, you're welcome (I say "rather useless" because I use a Core i7, and in my experience it's mostly useless under normal home-usage patterns).
Pairing them probably will end up in a hot system, but it should be perfectly safe: more probably you can't have utterly silence with such a setup, at least without running rather hot, but in case it would be your call.

Said that, the Big Shuriken 2 isn't an extremely proficient cooler: it's just one of the best results you can squeeze from a 58mm chunk of metal. More probably that not it's not up to the task to cool a Core i7 quietly under any circumstances (and so it should be for any of the other alternatives which have already been named).

Can the NT01-Pro (with a couple of slow running 80mm fans) be a way to do better?
Well, probably Silverstone didn't think the GD02 for benchmarking or testing (you won't race at Le Mans with a stock Rolls Royce Phantom, I guess) so, whether you should deal with synthetic loads such as "Prime95" or any Linpack based test, IMHO the NT01-Pro won't be able to do better.
On the other hand, in a real life scenario, there are much more chances to work out a perfectly usable rig, which can run enough cool and maybe enough quiet: from the test I read about that cooler, it isn't a stellar performer, but heat-wise it could work slightly better than most of the 60-70mm coolers around.

Summarizing, if you give up the demand to cool that CPU silently with all kinds of load, in real life the only serious drawback I see is that it's a rather expensive option: but it's expensive the GD02 too, isn't it? Have a nice time.

I think your test would yield different results with a tight fitting duct from the fan surrounding the heatsink down to the CPU. I understand this would not be a simple task but probably necessary to optimize air flow with a reversed fan. Thank you for taking the time to do your basic test & sharing your results.

It's irrelevant but I purchased the GD02 case years ago as "open box" very much reduced in price. This computer is for business use at home for engineering work so the i7 will indeed be put to good use, not wasted.

Well, curiosity just moves and improves the human kind, so whether you'll want to make some experiment you're welcome, and I'll be glad to see the relevant results.

Just about heat transfer, given that forced convection is more effective than natural convection, blowing air towards an hot object move more fluid (air) around it, opening a broader pathway to heat, due to the higher supplied kinetic energy (with the same effort, i.e. rotational speed): that's why the fan blows and don't sucks, and thermodynamics says to us that no ducting structure may alter that natural fact, and especially inside an ATX enclosure.

That's EXACTLY what a properly constructed duct would accomplish. I would argue likely more effectively as well since it wouldn't be fighting natural convection. There is why tower cases don't try to force airflow downwards.

I would agree that it is often MUCH simpler to direct air at the motherboard in most cases, especially small ones. However, once small enough, the case itself can become much like a duct so careful consideration to air flow alone may be enough.

The NT06-PRO is interesting but probably won't fit. Going to try NT01-PRO & I'll share my results here. It will be some time from now though.

Thank you for the useful discussion.

A duct would stop air from coming in from the sides, but it would still probably be less effective. Only a very few coolers are constructed with the intention of having a fan sucking air from it rather than blowing towards it. I think most fans perform better with the resistance being in front of the fan rather than behind it. It might be possible to construct fans that sucks better than the common fan, I dunno.
Natural convection is not very strong, otherwise we see a lot of setups with fans blowing up, to make use of it.

Yes the duct would have to be constructed to allow air to enter from the very bottom. I agree it may still not cool as proficiently but much much closer. The issue is that the CPU is designed to dissipate heat on the side facing up. That side always has a heatsink contacting it so direct flow onto that always helps. That is indeed difficult to achieve with a duct. Any such duct would also be quite involved to construct & likely tedious to install.

The fan design shouldn't be relevant since the pressure would be equally changed on either side, just opposite magnitude.

All these issues certainly favor blowing onto the CPU for a general solution but each case is different & other solutions can likely be adequate & perhaps better for other reasons (access to MB, for example).

Well, definitely I don't understand where you're getting at: more probably that not, it's due to my not so good english.

Eventually I just add that the natural convection you're describing looks like a (necessarily simplified) abstraction, which - in case - applies just to desktop laying enclosures: with reference to all the other kinds (tower, blade, cube, and so on) your assumptions on the expected contrast between the forced convection brought by a top-down blower and the natural convection just do not seem to apply.

Again, my best wishes for your experiments, I hope to read about the relevant results here on SPCR.

I read some on heatpipe design &, if built properly, I would expect extra airflow on the mating CPU heatsink to be less relevant. No doubt it's hard to imagine it not helping in most cases.

Finally had time to do more searching & found many other interesting options. I had no idea how many CPU cooler options were available these days! Many more than last time I built. The Scythe Kozuti is interesting to me. Probably noisy when pushed but maybe some well-placed acoustic material could help. I will also experiment with blocking case vents & interior air flow when I have time.

Does anyone have information on the stock cooler performance for reference? I will certainly do my own testing too, before swapping, but I'm away from that computer for a few more weeks yet & wish to order parts for when I'm working on it again. These discussions help. Thanks!