SPCR

some errata:

pg 1, spec table: "Motherboard to Fin: 36 + 8 = 44 mm 46 + 8=54 mm"
Might be clearer to say something like "side and rear clearance: 36 + 8 = 44mm; front clearance: 46 + 8 = 54mm"..then again, not sure if that's clearer

I love these surface-area monster coolers. I still use an Orochi or two, if you remember those mammoth coolers with wide fin spacing. The passive-cooling chart is fascinating.

Thanks Lawrence for the review.

A little disappointed into not performing significantly better than the HR02, i though the more area would make a bigger difference. And that U14S is a better pick imo being cheaper and having a fan included, which would mean that the new flagship from noctua or the new iteration of the NH-D14... should be a lot better.

Interesting stuff, as always. For the next time, I really hope you can try to use these kinds of heatsinks in a totally passive build. This would enlighten us about the current state of passive build temperatures. (Yes you did a fanless heatsink test in this review, but I saw that you had left the system fans on as well.) Readers could see at what temperature-expense the reduced db reading comes. Also, you didn't legitimize why you ran primemark at half-load.

Also, I was wondering; why no build on the recent 1150 based processor?
And, although the lapped cpu is a nice twist to the story, it is basically just that - many users will just use a nonlapped cpu and rather not mess around with its IHS. If one would lap his/her cpu ihs, then certainly it would be no big deal to lap the heatsink base as well. You didn't inform us about your thoughts on what the effect of this would be, temp wise/db wise.

Another thing is: does the heatsink block the first x16 pcie slot? It seems so big, and as I have seen it mounted on a M-ITX board, it seemed to block the PCIe board. That is something to take into consideration when thinking of buying this heatsink. Any thoughts on this?

Furthermore, I couldn't get the mp3 to download - the page didn't exist when i clicked on its link.

All in all a very in depth review. I really enjoy reading these articles (for years now)

These >1000 g coolers are simply insane. It's not just the diminishing returns aspect that bothers me, but it's a safety issue when you're moving the PC around (and possibly forget you have this kind of cooler in there) - but I guess there's a demand since they are still being made, even when a 400-600 g cooler can do a satisfactory or even superb job (NH-U12P for instance is 600 g sans fan, the TR TRUE Spirit 120 M is 470 g, both very successful in SPCR testing).

Das_Saunamies wrote:These >1000 g coolers are simply insane.

+1

I'd worry about the torque on the motherboard.

vatan007 wrote:Interesting stuff, as always.
1. For the next time, I really hope you can try to use these kinds of heatsinks in a totally passive build. This would enlighten us about the current state of passive build temperatures. (Yes you did a fanless heatsink test in this review, but I saw that you had left the system fans on as well.) Readers could see at what temperature-expense the reduced db reading comes. Also, you didn't legitimize why you ran primemark at half-load.

2. Also, I was wondering; why no build on the recent 1150 based processor?
3. And, although the lapped cpu is a nice twist to the story, it is basically just that - many users will just use a nonlapped cpu and rather not mess around with its IHS. If one would lap his/her cpu ihs, then certainly it would be no big deal to lap the heatsink base as well. You didn't inform us about your thoughts on what the effect of this would be, temp wise/db wise.

4. Another thing is: does the heatsink block the first x16 pcie slot? It seems so big, and as I have seen it mounted on a M-ITX board, it seemed to block the PCIe board. That is something to take into consideration when thinking of buying this heatsink. Any thoughts on this?

5. Furthermore, I couldn't get the mp3 to download - the page didn't exist when i clicked on its link.

6. All in all a very in depth review. I really enjoy reading these articles (for years now)

1. Fans in the system? You misread: there is no fan for those passive tests. P95 heats up an Intel CPU like no other real app; even at half load, it's obviously enough to overload any passive heatsink.
2. We always test on one of two reference systems, otherwise there's no way to compare heatsink results. 1150 CPUs are no challenge at all anyway, for any of the big high ranked coolers tested on the 1366 platform.
3. I think it's safe to say everyone reading these articles knows a lapped CPU & lapped HS represent the best possible combo -- assumingg the lapping is done well. We didn't feel any commentary is needed. The only reason our CPU was lapped was because it seemed to become too concave due to multiple installations/tests of high pressure convex base heatsinks. Read the lapped cpu article pls.
4. No.
5. Will try to fix that today.
6. You're welcome.

vatan007 wrote:Another thing is: does the heatsink block the first x16 pcie slot? That is something to take into consideration when thinking of buying this heatsink. Any thoughts on this?

It shouldnt block the first PCIe slot, the HR02 macho is just like 2 or 3mm from the PCIe slot, now the HR22 is wider by 10mm so it should, but its asymmetrical design makes it have the same length to the left side of the CPU as the HR02, so is kinda like HR02 with more metal toward the upper side of the case.

It seems so big, and as I have seen it mounted on a M-ITX board, it seemed to block the PCIe board.

On a mini ITX mobo will be dependent on the position of the CPU socket and the mounting orientation, this varies a lot, if its right next to the PCIe slot it will block it mounted toward the IO, but it shouldn't if you mount it 90 degree rotated as its flush with the CPU. Asrock haswell mini itx have the CPU socket opposite to the PCIe slot, this shouldn't block it with the normal orientation, Asus haswell mini itx come with the CPU more centered and shouldn't have an issue mounting it either, here is a pic of it mounted asus mini itx mobo,



I'm going to get mine tomorrow, i'll installed it on the weekend, ill do a prime95 on the HR02 with the TY150 and disassemble it, install the HR22 with the same TY150 and run prime95 again, ill post my results to see if there is any gain.... kinda thinking there wont =(

In all honesty im still angry and disappointed, this is what thermalright is marketing as their new flagship, 10 years of experience bigger area and so on... to gain so little over the HR02 Macho that cost half, weights less and comes with a fan. I know this heatsink will get bashed by overclocking sites out of its price tag, as it was with the original HR02 (none macho version), but i really don't care about the huge price tag, but it needed to perform better, much better, it should have taking the crown on temps at least to justify the cost, its a fanless oriented cooler that dont beat the NoFAN95, and with a fan barely any gains on the HR02 macho, i don't see whats its strength.

Thanks for the review.

I just bought the HR-02 Macho (three days ago) and I’m very pleased with its performance, the only thing I don’t like is the crossbar/clamp it came with, it lacks the centre knob. In any case, it seems I did a good purchase; it was dirt cheap over at the local computer shop too.

Wow - just when you thought a heatsink couldn't get any bigger or more complex! That is impressive.

Typo on the top of the 2nd page: " 8 x 6 mm thick copper heatpipes"

Hello friends,

I used to post here under the nickname ''SpyderCat", in 2003 ...


I just finished this summer my new build; an Intel i5 4670K (Haswell) on Gigabyte Sniper M5 in a Fractal Design Define R4 case.
The graphics card is an ASUS GTX 660.

I used a fanless Thermalmalright HR02 ducted to the standard 140 mm Fractal Design casefan @ 5 volts.
This rig is 24-hour-Prime95-stable @ 4.0 GHz, and I can say: " It's almost silent. "



I think you could get better results than in the test by:

* 1. Use "Countercurrent Flow Cooling" by rotating the heatsink 180 degrees, AND DUCT IT TO THE CASE FAN !, thus optimal benefiting from the asymmetric design.

* 2. Close the sides of the heatsink in order to force the air to travel 'the long way'.


If I had to build a PC now I would use the Thermalright HR-22, fanless and ducted.


Regards,
Han.

If i would buy a 100$ cooler, i expect it to be perfect right out of the box.

Besides, one of the few coolers that are more expensive than the CPU it might cool fanless.

I often wonder if the heatsinks put out by these companies to run without a fan are done so assuming a non maximum load. By that I mean what most would consider "normal computing", without periods of high cpu usage that run minutes long. I know that this is far from the worse case scenario, and tests like Prime 95 are intended to stress the system to represent something close to it. Personally, I usually just run a fan on the cpu cooler at very low rpm and it usually fine for what I do.

That m-itx photo is hilarious, Abula.

The weight of massive coolers is a concern, at least when the mounting design is inadequate; thankfully, the push-pin mounting era seems over. Since most PCs only get one CPU heatsink, a cooler made large enough to gain insignificant returns can be worthwhile over many years. "Silent" isn't the same as "quiet", so people here tend to consider even minimal improvements.

For all the danger of weight and mounting system, and for the high price of 100 dollars, this thing needed to best all air coolers. It didn't, not in any situation or setting. A good cooler, to be sure, but all things considered, it's a sub-optimal purchase.

My HR22 just arrived, its beautiful, even better looking than the original HR02, crome looking on the top like the newer Silver Arrow SBE, but i wished it perform as well as it looks, but its not the case, my experience is similar to SPCR, the cooler performs way to close to the HR02, and at double the price its simply not worth it. I'll leave a couple of runs of prime95 blend test on both cooler, both done today with very similar ambient temps, under the same case with the same fans and settings, nothing really changed aside from swaping the HR02 for the HR22.



For people looking for coolers, the Thermalright HR02 Macho and Scythe Mugen 4 are still the best air coolers for the money, and for people looking for higher end cooler avoid the HR22, look into the twin towers like Noctua NH-D14, Thermalright Silver Arrow SBE, Phantek PH-TC14PE, Prolimatech Genesis.

Thanks for that. I think that the best fan for the HR-22 would actually be the Noctua NF-F12. That's because despite only being a 120mm fan its Focused Flow feature would make a real difference with this particular cooler. The airflow would be concentrated in the central section of the cooler where the heatpipes are and this concentration of air would I suspect cope much better with the sheer depth of the HR-22. There is a YouTube video which illustrates Focused Flow rather neatly.

For anyone interested, there is a Black MACHO version coming (not related to the HR22), Macho Black. Probably wont be as cheap though, as 2 fans and black, i like the looks.

Another review just came out, Hardwareluxx Thermalright HR-22 in the test

Sadly even on passive gets beaten out from Phantek PH-TC14PE, which is a cheaper heatsink by at least $20 =(

In passive mode, ie without active fan on the radiator, placed the Thermalright HR-22 with an average temperature of 58.75 ° C on the second space of the test field and succeeds in playing to operate the Xeon E5 2678W within its thermal specifications. Only the Phanteks PH-TC14PE can still just over one degree set ° C ahead of the HR-22.

That one degree sure is critical.

I have been using Scythe Mugen 3 CPU coolers on my Xeons for the past four years which have been solid performers. Now I want to switch to new coolers for certain reasons and my main contenders are various Thermalright Macho versions (HR-02, grand macho and HR-22) and Scythe Ninja 4. For me, quiet cooling has the uppermost importance (by which I mean a low rpm fan setting but not totally fanless). According to SPCR, Ninja 4 is now the champion of quiet cooling, while larger Thermalright bigger coolers such as HR-22 barely offer any advantage if any over their smaller brother, the HR-02 or Scythe Ninja 4. Other reviews on the other hand, shows some advantage of advantage for these super-sized heatsinks over HR-02. According to
http://www.xtremesystems.org/forums/sho ... d-Macho-RT,
at a low 600rpm fan speed, there is indeed a significant advantage anywhere from 6C to 11C depending on the cpu's clock speed. This is hard to overlook. I know test methodologies differ but still wondering what causes such different outcomes. Right now, prices on Ebay are not much different while Ninja 4 is still somewhat less expensive as well as having the advantage of its symmetric design yet the disadvantage of potential conflict with RAM installation. So the choice remains a moderate puzzle for the moment.