FSP hits the heatsink market with its Windale CPU coolers

FSP is mainly known for making power supplies for PCs, consumer electronics, and medical and industrial equipment. Someone in the company has apparently decided the time has come to start adding heatsinks to other spots in PCs. The FSP Group's Windale 4 and Windale 6 pair of CPU coolers is the company's first foray into the PC cooling market. The company seems to be placing the Windale units as alternatives to some of the most popular cooler offerings.

We'll start with the big gun—in this case, the Windale 6 and its arsenal of six 6-mm direct-contact heatpipes. The copper tubes connect to a large array of black-anodized aluminum fins. A single 120-mm fan with blue LED illumination and a rubber-isolated mount whispers sweet nothings to the big fin stack. FSP says the Windale 6's performance compares favorably to Cooler Master's Hyper 612, which is also built around a six-pack of heat pipes. The Windale 6 weighs a hefty 29 oz (823 g) and measures 4.8" x 4.3" x 6.5" (or 12.2 cm x 11.0 cm x 16.5 cm).

As clever gerbils might figure out from the name, the Windale 4 pares the number of heatpipes down to four. The black anodizing and the blue LED in the fan are also lost in translation from "range-topper" down to "midrange cooler." Fortron specifically compares the Windale 4 to competing coolers from Cooler Master, Be Quiet!, and Thermaltake, including the ever-popular Cooler Master Hyper 212. The Windale 4 is more reasonably-sized than its bigger brother, measuring 4.8" x 3.3" x 6.2" (12.2 cm x 8.3 cm x 15.8 cm) and weighing in at 21.9 oz. (620 g).

Both coolers list support for Intel LGA 775, 1150-1156, and 2011 sockets. AMD users can also use the Windale coolers, which support the red team's myriad AM and FM sockets, including Ryzen's AM4 socket. Those using AMD's decidedly low-budget AM1 platform will have to look elsewhere, though.

FSP didn't offer pricing or availability information for the Windale 4 and Windale 6. Neither cooler uses the offset fin stack design we've seen in recent coolers from competitors, so buyers will need to double-check the clearance for memory height before buying.

Comments closed
    • albundy
    • 3 years ago

    my $9 zalman cnps10x tower cooler still does the job.

    • Captain Ned
    • 3 years ago

    And this beats my 10 year old but still excellent Thermalright Ultra 120 Extreme just how?

      • DPete27
      • 3 years ago

      Yeah, seems like an awful saturated market to dip your toes into.

    • tipoo
    • 3 years ago

    Did they pick names for maximum brainfarts? Took me a while to parse the title in the context of this site with “FSP” – fps and Windale – Windows 😛

      • derFunkenstein
      • 3 years ago

      FSP is like Fortron, Sparkle, and something else. I forget exactly and Google isn’t helpful.

        • continuum
        • 3 years ago

        Fortron-Source/Sparkle Power International. Hence they dug around and got FSP.

    • derFunkenstein
    • 3 years ago

    I like the six-heatpipe design, but man, the bases of those things look rough.

      • morphine
      • 3 years ago

      I’ll be honest: I never understood why people care so much about a mirror-finished base when you’re just going to stick thermal paste over it. Assuming there’s even a difference, it’s bound to be so minimal as not to matter at all.

        • derFunkenstein
        • 3 years ago

        I don’t necessarily see the benefit of a mirror finish, but there is a middle ground. I’m concerned with pits where air gets trapped and doesn’t transfer heat as well.

          • Chrispy_
          • 3 years ago

          I’m pretty sure that the advantages of an exposed/flattened heatpipe outweigh the disadvantages of putting a machined baseplate on them.

          Sure, there’s a small gap between adjacent heatpipes that will fill with paste but the IHS on your processor is a bigger barrier to heat transfer than the tiny wedge of paste.

          As for the difference in surface finishes affecting heat transfer, nope. That argument is irrelevant (within reason) unless you have de-lidded your chip. The IHS itself is rough enough (on both sides) that it is by far the more relevant issue (and it really isn’t an issue at all, compared to quality of the TIM between the IHS and the die!

          So if the TIM inside the CPU package has 10x more significance than the IHS, and the (lack of) IHS flatness is 10x more significant than the surface smoothness of the IHS, the paste and the cooler finish are basically insignificant. Worrying about it is like sanding and varnishing the front door when the house is on fire and there’s an earthquake destroying the whole neighbourhood.

        • tipoo
        • 3 years ago

        There’s mirror finished and then there’s first round in Forged in Fire finished, this leans towards the latter 😛
        Particularly this has seemingly significant air gaps between each heat pipe, that would be a mountain in the world of thermal paste.

        Maybe they tested it and found it doesn’t matter, but it’s like the inside of a Macbook or something, the details show care.

          • Welch
          • 3 years ago

          It supposedly doesn’t matter and in fact helps the surface to have a somewhat “rougher” surface, assuming a few things.

          If a surface was mirrored then it has near a perfect flat surface, end of story right? However, if the surface has small ridges and valleys, assuming they can be filled with thermally transferable material greater than or at least equal to the metal itself… Then you’ve essentially created a larger surface area by which thermal energy to be dissapated.

          This is my understanding after years of seeking a perfectly lapped heatsink only to have finally found a source claiming the above, which makes sense when thought about.

          If anyone knows otherwise and has a great link, by all means share.

            • tipoo
            • 3 years ago

            “However, if the surface has small ridges and valleys, assuming they can be filled with thermally transferable material greater than or at least equal to the metal itself.”

            This is where the assumption falls – thermal compound does not transfer as well as pure metal, just a lot better than air. The most efficient transfer would be a heatsink soldered right to the CPU in theory, but thermal compound is a necessary compromise for swappability.

            Proof of this: Soldered integrated heat spreaders vs heat spreaders with thermal compound inside, the second thermal compound layer adds a few degrees, sometimes significantly

            So with that in mind, the forgone conclusion is “as little thermal paste as possible to fill all air gaps”, while an uneven surface, while increasing surface area, would create larger thermal compound gaps.

            It is, in the end, a lot of thinking over a few degrees though, i think people used to care more a number of years back (when the advice was to spread a fine layer with a flat credit card), now it’s more “this is good enough to not make a difference” with the blob method.

        • UberGerbil
        • 3 years ago

        The thermal paste is not as thermally conductive as pure metal-to-metal contact. It’s just much better than an air gap. So it follows that the closer you can get the two surfaces to perfectly flat, the more metal-to-metal contact you have and the less thermal paste you need, and the better the heat transfer overall. (In an ideal world, you could mate them in a vacuum and they’d spontaneously [url=https://en.wikipedia.org/wiki/Cold_welding<]cold weld[/url<] themselves together). A perfectly flat metal surface polished to the point it has no irregularities is going to look pretty much like a mirror. Certainly if you can see any bumps or distortions your heat transfer is going to be less than optimal. Whether that actually matters depends on the situation. For overclockers who are taking the CPU to the outer limits and need to get as many watts of heat away from it as fast as possible, I can understand the obsession. For most people, a degree or two different, meh.

          • synthtel2
          • 3 years ago

          As far as distortions, a slightly convex base is purported to be helpful, presumably because it results in a thinner TIM layer over the die.

      • Firestarter
      • 3 years ago

      I wonder how much the outer heatpipes even do, considering how small modern CPU dies are

Pin It on Pinterest

Share This