Designed for Socket A exclusively, the Alpha PAL8045 (Dan's reviews are known to give heatsinks a much higher C/W than usual, but he's very accurate. He considers the PAL8045 to be 0.53 C/W) was regarded as the king of all Socket A heatsink coolers, though in its later years upstarts such as the Swiftech MCX 462 and the various all-copper Evercools managed to supplant it - But only with the help of a screamingly loud fan. As far as quiet, efficient heatsinks go, the PAL8045 was not beaten until the heatpiped tower coolers started to appear five years later, such as the Coolermaster further down this page.
The PAL8045 was so efficient that increasing the fan to one with four times the output made very little difference to the effectiveness of the heatsink. Using Alpha's "Microforge" process, the aluminium cooler had, as can be seen, numerous small hexagonal pins instead of the traditional fins or vanes. This gave it a massive surface area and increased its density, both important for a heatsink. This caused a large pressure drop from the top of the heatsink to the bottom which is normally a bad thing. Air couldn't flow as freely through the PAL8045 as it could a traditional cheap fin thing (see the Evercool below for a typical example). This meant the PAL8045 didn't benefit much from very powerful fans, but the trade-off which Alpha hit a perfect sweetspot on was that the PAL8045's surface area was massive. The heatsink didn't have the same kind of airflow as hairdryer emulators but it made much better use of the airflow it did get.
The copper inlay was used for its high thermal conductivity, it spread the heat load across the base of the heatsink so that the outer areas were more effective and the heatsink could be made larger than the typical 60mm. In this case, the heatsink was a full 80mm and one of the first of the 80mm heatsinks which used fans which before then had typically been reserved for case cooling. 80mm fans are able to move more air at a lower RPM rate, so be just as effective at a lower noise level.
Being a bolt-through-holes cooler, the heatsink required a motherboard compliant with AMD's mounting hole specifications which not all of them were. Common designs from Gigabyte, Asus, Abit and DFI were compliant, I used it with an Abit KT7, KT7A, KX7-333 and Asus A7N8X-X, but cheaper boards typically didn't bother - Check the PC Chips/ECS board in the motherboards section, the two mounting holes comply with no known specification.
It cost a fortune, it wouldn't fit all motherboards, it required assembly, it usually didn't come with a fan, but if you could fit it, use it and set it right...oh boy were you in for a treat.
There are two ways of making a heatsink. You can do the smart way, design a precision made instrument (see the Alpha above) and charge a fortune for it but sit back and know you're putting the newbies in their place. Or you can put together a few fins and a screamingly powerful fan on top (Anyone ever seen an OCZ Dominator? No? Bet you heard it coming, though!) of an otherwise unimpressive heatsink and be guaranteed great performance. Sure, you wouldn't beat the Alphas of the world, but you'd be about half their price. The big tall fin heatsinks, like this one, responded very well to an awful lot of airflow.
With a screamingly loud Delta 70mm fan and an all-copper design, the Evercool CUD-725 belongs in the latter category. It's actually heavier than the Alpha PAL8045 (though not by much) but a lot cheaper to make. It's a skived-fin copper thing completely lacking in grace and subtlety. That's not to say it was bad, far from it, but that it wasn't excellent. It performed within a hair of the Alpha PAL8045, but at a much higher noise level.
A simple design, done well, the rulebook followed to the letter. Most testing placed the CUD-725 around the 0.60 C/W level.
Coolermaster Hyper 6+
Sat atop an ATX PSU for scale (as are most of the other heatsinks on this page), the Coolermaster Hyper 6+ is a six-heatpipe two-stage tower cooler with copper heatpipes and aluminium vanes. In the original Hyper6, the vanes were also copper which added a great deal to the mass of this behemoth, but the aluminium version weighs in at 'only' 971g.
Many people will tell you the Hyper 212 was Coolermaster's first great tower heatpipe cooler, but the Hyper 6 was. The 212 was a cost reduced version of the Hyper 6, the latter being Coolermaster's first big tower cooler. Coolermaster had introduced the heatpipe to the mainstream with coolers like the HHC-001 which could use much larger copper vanes than anyone else, as it had two heatpipes to lift heat to the upper parts of the vanes.
In use, the Hyper6+ was able to keep an estimated 110W load (2.8GHz dual core Opteron at 1.45V) at 65C with the fan turned down to only 80% PWM - From this I work out an excellent real-world thermal resistance of 0.20 - 0.25 C/W. In design, it has a large copper slug on the bottom to make contact and house the hot end of the heatpipes but where most coolers then have a whole load of open space doing nothing, the Hyper6+ fits a first-stage heatsink which cannot harm performance but probably won't make much difference. The aluminium vanes are closely spaced but parallel, well fitted and offer very little resistance to airflow. It's big, effective, imposing, weighs almost a kilogram...and the fan glows blue. What more could one want?
Alas, the universal fitting of the Hyper 6+ fits many sockets with mounting holes...except Socket F and Socket AM2(+). When I moved to AM2, the Hyper 6+ had had to be retired...Until, like with most of my hardware, it found use in this server. Quiet, extremely high performing and reliable. Just what you need in an always-on system.
Other than the lack of AM2 compatibility, the other problem with the Hyper 6+ is its use of a non-standard 100mm fan. 92mm and 120mm are standard, but 100mm fans are extremely difficult, almost impossible, to find.
Though i did not try this myself, it seems the Hyper 212 LED AM4 upgrade kit would actually fit a Hyper 6.
Corsair H80 CLC|
CLCs came of age with the advent of very small high pressure pumps and the use of propylene glycol as a coolant. Very quickly models from Fractal Design, Corsair, NZXT, Swiftech, etc. all appeared at very attractive price points. Even AMD got in on the action and began fitting them from the factory to some high end GPUs. Mostly, they were identical and differed in the fans supplied, because they were being made by the same supplier: Asetek. Cooler Master, Apaltek and some others got in on it, but even they largely used the same radiators.
Corsair's H80 runs three profiles, cycleable via a button on the CPU plate, predictably "low", "medium" and "high". The pump, in my case, is silent, but others have reported noise. The fans do not spin down much at all when on medium and high, and always make noise. On high, quite a lot of it, so those who can thermally control their fans are advised to put the pump on full and control the fans themselves.
It is immensely effective. It sustained a full load highly overclocked Phenom II X4 955 BE at 3.8 GHz, which drew an estimated 220 watts, at below 70C.
The advantage of a CLC, indeed any liquid cooler, is that the heatsink or radiator device can be placed out of the way, it's not limited by whatever will fit mounted to the motherboard. This means it can be very big, very heavy and very dense, all highly desirable traits for a heatsink.
Of course, an overclocked Phenom II won't last anyone forever, and it was replaced by a Core i5 2500. The H80 was set on "low", and controlled the fans itself.
Under full load from Prime95 small-FFTs, the Sandy Bridge reported itself as using 76 watts total. This is hardly going to strain anything, so OCCT was brought in. The TDP's 95 watts, let's try to get closer to it? OCCT managed 62 watts and the CPU didn't even break 60C. This became a challenge! Manually limiting OCCT to 4 threads and a small data set got package power up to 76 watts and temperature to 65C, but surely we can do better?
OCCT's AVX capable Linpack came next. AVX is responsible for a lot of power, and we hit 80 watts quite quickly while temperatures stabilised at 70C. I can only assume the thermal interface is rubbish or the H80 needs more pump speed - so it was set to "Medium" while the test was still running. Fans ramped up, pump RPMs (reported via a 3 pin fan connector, a neat touch) ramped up, noise ramped up and... Nothing. Temperature remained flat at 70C, a few flicks to 71C, then power dropped to 48 watts before resuming at 80 watts, ruining the nice flat temperature trace. The temperature made no further threats to reach 70C. Clearly something more reliable a load than OCCT was needed.
IntelBurnTest got to 78 watts but was all over the map. 7Zip's benchmark got to 46 watts and was a bit of a joke, but it's mostly memory bound anyway, and actually scored worse than the peak result I'd seen on the Phenom II X4. Tweaking it to fit the dictionary in L3 cache and running a few more threads helped close in on 50 watts, but this isn't a stress test for thermals. An old favourite was "burnk7", hard to find, and designed to present maximum load to an Athlon back in the day. It was also a very high load for Core and Core 2 processors, but evidently not for Sandy Bridge: Which hit 51 watts only. I then updated Prime96 from 27.9 to 29.4, which is AVX aware. CPU hit 75 watts immediately but went little further, stabilising around 76.5W. I left this going at 64C and ramped the H80 up to "High". Temperature fell slowly to 62C and stayed there. With a resounding "Meh" from this whole round of testing, I set the CLC back to its "Low" profile. Temps very slowly rose back up to around 66C.
The only conclusion is that a low 70-80 watt load just isn't enough to let the Corsair H80 strut its stuff properly.
AMD/Coolermaster Wraith Stealth|
If you bought a Ryzen 5 or a Ryzen 3, chances are the included stock cooler was this thing. There were several versions of it, some with more powerful fans than others. This one came with a 2,600 RPM fan, but all are variants on Coolermaster's "Standard Cooler" series, completely no-nonsense radial fin aluminium things. AMD got a little surround clipped (and screwed, for some reason) onto the fan.
Before we talk about this actual cooler, let's put some mythology where it belongs.
Myth: AMD uses multiple sources for its retail or PiB coolers.
Truth: AMD uses CoolerMaster exclusively.
Myth: AMD designs the coolers, CoolerMaster manufactures them
Truth: They're variants (semi-customs, if you will!) of existing CoolerMaster heatsinks. None are precise, but they are close enough as to be variants on the same basic design.
Wraith Stealth: Standard Cooler I30
Wraith Spire: Standard Cooler I70
Wraith Spire (old): Standard Cooler A71C
Wraith Max: Vortex Plus
Plus, why would AMD get into the HSF design business? AMD has been divesting its capability: It cannot even lay out its own chips! In addition to die layout and synthesis (designs sent to the fab for manufacture) AMD also cannot produce its own chipsets anymore, it licensed AsMedia designs rather than update its existing ATI-inherited and extremely obsolete chipset series. Why, oh why, would it then decide to design heatsinks?!
CoolerMaster is an excellent vendor to partner with, they have a good reputation, a very wide product range, from vertical fin coolers, to radials, to tower stacks, to closed loop liquid coolers.
Like vertical fin heatsinks of yore, this is the "Keep it simple, stupid" of heatsinks. If we change the design with the same materials, we make it worse. It's why both Intel and AMD settled on this design
The radial fin things supplanted vertical fin heatsinks in the late 2000s not due to performance, but due to practicality. They worked just as well in less vertical height and spilled air in all directions, so giving airflow to nearby VRMs and RAM. Coolers like the AVC Sunflower and the various ThermalTake "orb" coolers (everyone had a TT BlueOrb on their chipset at one point!) pioneered the design, with a copper core and a spiral of aluminium vanes around it, they ran very, very well. They allowed a larger heatsink, so a larger fan, so a quieter fan moving the same quantity of air.
The original Wraith cooler was very highly regarded, and the Wraith Spire was also extremely effective, so what of the Wraith Stealth? Small, low profile, and quiet are never good signs for cooling effectiveness.
A completely standard running Ryzen 5 5600X has a TDP configured to be 65 watts. The Wraith Stealth hits thermal limits on this, at 95C, causing the CPU to back off an all-core load to around 4.1 GHz. CPU package power (PPT) hovered around 72 watts during a Cinebench R20 run, being thermally limited. This was actually better than the much larger and more impressive looking Nova Pulse RGB cooler.
It's not bad, not good, recommended for loads of 65 watts or lower. The Ryzen 5 3600, at 65 watts, is about fair for this cooler. However, the Ryzen 5 5600X, while defaulting to 65 watts, is much, much happier with more headroom available. The Wraith Stealth has no headroom at all.
Precision Boost Overdrive can only just tickle the 4.65 GHz limit of the Ryzen 5 5600X with the Wraith Stealth fitted. People wishing to let PBO push performance higher are strongly advised to go with an aftermarket cooler intended for 90-100 watt or higher loads, something like the Coolermaster Hyper 212 (series) or better. The Wraith Stealth in my estimation shouldn't be used for loads higher than 50 watts.
Nova Pulse RGB|
Looking at this, it looks big, it has a lot of surface area, big fan, it can't be too bad, right? A 75 watt load caused a temperature rise of 60 degrees above ambient, causing the CPU to thermally throttle down to 60 watts, at which point the temperature stabilised at 85C, 60 degrees above the ambient of 25C.
That is awful!
The cooler looks cool enough, and it's a big imposing thing, but it only looks big. Most of the interior is empty space. Where AMD's Wraith Stealth (above) has a solid aluminium core to sink heat into, this has nothing.
Pictured: An objectively bad heatsink
Why did I buy it? Two reasons: It looked a lot better than it actually is, not knowing the damn thing is hollow, and it advertised Intel LGA2011 compatibility, something I needed for a Xeon E5 2640. Both these things turned out to be false, but I did manage to bodge a good fitting to LGA2011 by cutting the supplied universal ring adapter to fit and using spring-loaded bolts from another heatsink.
It performed terribly on the Xeon, but this was put down to it being fed pre-warmed air from the first CPU in the system (Dell Precision T5600 dual Xeon workstation).
I then modified an actual LGA2011 heatsink to fit the limited clearance of the Precision chassis (by crimping and bending heatpipes) which worked far, far better. The Nova Pulse RGB was then taken to a Ryzen 5 5600X, to replace the stock Wraith Stealth.
It performed worse than the Wraith Stealth. Significantly worse. It not only ran hotter, at lower peak clocks, and lower performance, but it was also noisier!
This is a heatsink for loads below 40 watts, it is larger than it needs to be and noisier than it needs to be for that kind of load. It is proof that we can make a heatsink which is large, has large vanes, and pushes a lot of air, but doesn't work. Primarily, the mass of metal close enough to the base is far too low, and the vanes are both too long (the edges of the vanes do not warm at all) and too thin. With just one heatpipe from the base and circling the upper edge, this would be a very much more effective design: Zalman's VF900 GPU cooler, for example, which I had on a GeForce 7900GTO.
CoolerMaster Hyper 212 Black|
Some folk will tell you that the Hyper 212 was CoolerMaster's first successful heatpipe tower cooler. Those folk have never used the Hyper 6 or Hyper 6+, both of which are immediately ancestral to the original Hyper 212. The Hyper 6 used six heatpipes, three on each side, for very powerful cooling performance. It was able to handle a CPU topping 200 watts.
The largest production cost on a heatsink is the heatpipe array, so in making the Hyper 212, CoolerMaster used longer heatpipes and fewer of them: Four. Each heatpipe goes up both sides of the tower stack, and is therefore more effective than one heatpipe on the Hyper 6. However, the Hyper 6 used two heatpipes where the Hyper 212 used only one.
This then has what seems to be four heatpipes on either side against three heatpipes on either side, yet the Hyper 6+ performed just as well as the larger Hyper 212 did, as each heatpipe on the Hyper 212 is less effective. Why not just continue using the Hyper 6+? Well, the Hyper 6+ had to be fitted using an improvised clip going through the fins of the small bottom heatsink, and it is a big, heavy cooler. A CPU mounting lug, particularly just one each side, is not rated for that kind of mass.
Comparing to the Hyper 6+ was perhaps unfair: The Hyper 6+ was positioned at a significantly higher market point, where the Hyper 212 series is intended as entry level tower coolers, but still full tower coolers. The half-tower coolers sometimes seen ae significantly worse.
The Hyper 212 Black was extremely cost effective, as was the point of it, and found itself used with an AMD Ryzen 5 5600X. Performance results are pending the 5600X being RMAed, but expected to never pass 80C under maximum possible load.
The Early Days, 1990-1995|
As with all things, heatsinks had trends. The first PC chip coolers came in the early 1990s as CPUs like the Motorola 68040 and Intel 80486, as well as some RISC CPUs, started using more than the five watts a ceramic chip substrate could handle. Initially, heatsinks were small things, barely 10 mm high, and usually 40-50 mm square. These got us up to 10 watts, and when we added small 40 mm fans on them, as much as 20 watts.
Scale was very different back then. We derided the original Intel Pentium for being power hungry, hot, and (as it turned out) unreliable. Right on the ragged edge of technology, they drew around 16 watts: Triple the power of anything else at the time. They came in a ceramic package, the CPU die actually under the package on the pin side of the PGA and heatsink fitting was crude at best. Flip-chips were still some time in the future. Intel's own heatsink was a many-pinned black anodised aluminium square which didn't have a fan. For the most part, only the PC's PSU had a fan, which was exhausting. This was historical, PSUs had just changed to being switch mode after an age of being hot, inefficient transformers.
This brought us to 1995, and the first mainstream Pentium processors. Socket 5, for Pentiums, had two heatsink lugs on the sides, to clip a heatsink to. Technically, it had two on each side with rotational symmetry, but the two offset ones were poorly used.
By the transition of Socket 5 to Socket 7, all heatsinks were small, no more than the size of the CPU substrate and thermal performance was an afterthought. A Pentium-MMX 200 MHz might run at 40-50C. 50C was considered quite warm. The 50x50 mm heatsinks did grow small fans, however.
At this point in history, some famous names in cooling began to appear. The market was basically square finned (or pinned) heatsinks, 60x60 mm square, which clipped onto the two mounting lugs. Zinc oxide thermal grease or paste was universal, but graphite pads were known. So, when larger heatsinks from thre likes of GlobalWin or ThermalTake appeared, people took notice. Custom PCs were beginning to take off, and who didn't want the unique ThermalTake BlueOrb chipset cooler (never necessary!) or the GoldenOrb CPU cooler (inferior, but attractive). ThermalTake made its name on these weird but cool looking coolers, and were among the first to use a flush-fit fan as a centrifugal blower.
As slot-based daughterboard CPUs came back into vogue, briefly, Slot-1 was already using a stronger through-hole mechanism and Slot-A was compatible. Intel's Socket 423 and 478 for Pentium 4 then went their own way.
I used a GlobalWin FOP 32-1 at this time, though soon replaced the fan with something more subtle. It replaced the FDP-32, and was in turn replaced by a PAL8045.
AMD's Socket A was the first variant on the standard two lugs by having two large lugs and two smaller, wider ones. Intel's Socket 370 also used the same heatsink fitting. With Socket 7, Socket A and Socket 370 all using the same fitting, this was the last time when a single heatsink mounting worked "across the board".
Issues were beginning already, however. AMD's heatsink retention pressure was absolutely massive, and caused many fitting accidents, crushed dies (we called them "cores" back then, the word did not have its modern meaning) and damaged motherboards. A cooler rated for Socket 370 or Socket 7 would fit a Socket A CPU, might even work, but it usually wouldn't work well.
Around the time of the Pentium 4, heatsinks were one of two categories: 60 mm square and with a calm fan, or 60 mm square and with a roaring 6,800 RPM monster. About this time, copper appeared as a heatsink material. First as inserts or bases, but also sometimes as entire skived heatsinks. Copper was a greatly superior material to aluminium due to its higher thermal conductivity.
By 2002, heavy based aluminium coolers were beginning to sport copper inlays to spread heat to outer pins or fins, allowing a larger heatsink. The Alpha PAL8045 was a perfect example. It could cost as much as the CPU it was cooling, but did it do an exceptional job. Some companies toyed with all-copper designs, which were crazily heavy but that bit more effective. Skived fins appeared, allowing an increased surface area. Early skived fin heatsinks had a fin density of around one per mm, but today that can be 5-10 fins per mm and is used to massively increase surface area in waterblocks. Extruded aluminium has a minimum feature size of around 2 mm, which limits fin size and spacing.
Heatsinks started to actually matter and CPUs had appeared north of 70 watts (the Pentium 4 1.5 GHz introductory model could peak at over 100 watts per its electrical specification), meaning heatsink design now mattered. A very large aluminium heatsink had the problem of the aluminium furthest from the base, so usually the top of the fins, were so far from the heat source that aluminium's thermal conductivity couldn't take heat to them.
|Material||Thermal Conductivity (W/m.K)|
While many told us that aluminium made a bad heatsink material, it was way better than most other metals. Copper beat aluminium by quite a lot, and silver was a little better than copper. Some heatsink makers did silver plating for... well, bling really. As a material, copper was both heavier and more expensive than aluminium. Industrial processes to mill aluminium were very mature, so a precision formed aluminium heatsink was cheap. Copper couldn't be formed the same way.
To make a copper heatsink, we'd either mill it using saws and grinders, or skive it with large hardened steel blades. The Evercool CUD-725 on this page is skived. Much larger than that, however, and heatsinks became too heavy for the motherboard to properly support.
AMD's Socket 939 and Intel's LGA775 were long-lived standards, with standard holes in the motherboard (AMD even provided a lug-attachment) and a backplate offering to spread the load of a heavy cooler, with threaded screw keepers poking through the motherboard's holes. This allowed heatsink innovation.
Skived fins had a short life. They increased pressure gradient, meaning they didn't much benefit from high airflows and, if too dense, didn't benefit from any airflow! The use of heatpipes gave us taller heatsinks. Even skived fin heatsinks could not have their fins too tall, heat conduction from base to tip, even in copper, was insufficient for a heatsink to be more than around 50 mm tall. Adding higher fins just didn't help: As any GlobalWin FDP-32 owner would tell you.
In the mid-2000s, a company called Zalman started to innovate with its "Computer Noise Prevention System" or CNPS heatsinks. They were a huge number of copper or aluminium fins and a very large, but slow, fan. Zalman brought heatpipes and large fans (first 92 mm fans, then 120s) together to prove that exceptional cooling performance did not need to be so noisy.
Heatpipes became low cost and mass market. Add a heatpipe soldered or compression fitted into the base, bent around and running through the upper portion of a heatsink's fins would allow a much larger heatsink. Indeed, it was possible to have the heatsink fins not making physical contact with the base at all, but instead suspended on heatpipes. This allowed the first "tower" coolers, two or three heatpipes bent at 90 degrees, then a stack of aluminium vanes through which a fan blowed. By 2006, anyone who was running top end CPU would be putting a heatpipe tower cooler on it, but it took until 2010 for them to go fully mainstream.
This started, or enabled, an arms race. Where we made frying-egg videos of a 68 watt AMD Athlon in 2001, joking how 68 watts was way too much, Intel had quietly specified Pentium 4's maximum power to north of 100 watts. By 2007, top end Core 2 Quad processors were running at 105 watt thermal design power and 140 watt absolute powers. AMD's Phenom X4s topped out with 125 watt and 140 watt models. At this point, overclocking (e.g. the legendary Q6600) could top 200 watts and a huge heatsink was needed for that. 120 mm fans, four, six, more heatpipes sprouted, with giants like the CoolerMaster Hyper series and the Scythe Ninja series. The heatpipes allowed the heatsink to be lifted above VRMs and RAM slots, so it could overhang them, and be truly massive.
The CLC came of age. A company known as Asetek developed a way to make very, very, very fine fins in copper cheaply. Five fins per mm meant a very small amount of copper could have enormous surface area. With a low volume, low profile liquid pump and a high boiling point coolant like propylene glycol, a "closed loop cooler" or "closed liquid cooler", a maintenance-free 120mm radiator, like the Corsair H80, offered exceptional performance. Many opined that heatpipe towers had finally met their match.
Liquid cooling, usually water with wetters and antimicrobials, has always had a major advantage over heatsinks: Radiators could be very, very, large. The first watercooling loops used aquarium pumps and motorcycle radiators to deliver incredible levels of performance, but they were always a very niche, very custom thing.
Most people were happy to ignore watercooling, but when CLCs arrived, they were extremely hard to ignore. Even basic CLCs mixed it with the very best air coolers. CLCs became larger, growing to two 120mm fans, then three. Cases with space for them became mainstream.