The Thin Mini ITX motherboard format is a great choice for an inexpensive, compact, and low power desktop PC. The motherboard format was created by Intel nearly ten years ago in 2011, primarily for All-In-One PCs and Small Form Factor PCs for the Sandy Bridge platform. The format has slowly declined in popularity since then. Although several Thin Mini-ITX motherboard have officially been announced for Intel’s latest 10th generation CPUs, as of September 2020 none of them have made their way to any major retailers yet. To my knowledge, only one Thin Mini-ITX motherboard has been developed for AMD’s Ryzen platform, ASRock’s A320TM-ITX. Thin Mini-ITX motherboards have certain benefits and drawbacks. They have onboard DC power conversion and generally accept a single 19-volt input, allowing relatively inexpensive laptop style AC-DC converters to be used. To keep their height down, they use smaller SODIMM memory modules instead of standard DIMM modules. They generally do not include any PCIe slots, but newer boards do include M.2 slots. If the M.2 slot sits right above a chipset though, I don’t actually recommend using it. Lastly, they do not have very robust power delivery components, so are best used with low power 35 W CPUs. Despite their drawbacks, they can be used as part of an inexpensive home theater PC or a simple home or office PC. The best fanless coolers are large and heavy to adequately cool power-hungry high-end CPUs and GPUs, but if our goal is to cool a low power CPU only with a tiny motherboard, we can get away with a very small integrated heatsink-case.
What is the best fanless Thin Mini-ITX case? I did a direct comparison of four Thin Mini-ITX heatsink-cases that I could get my hands on. Included were the Streacom FC2, Akasa Euler S, Akasa Galileo T, and Turemetal UP2. There are other versions of the Akasa Euler and Galileo, but these two are my favorites. So which one is the best? Well, the main concern with fanless coolers is their ability to keep the CPU from thermally throttling. I tested the four heatsink-cases with the same components at various power levels to see how each case does under the same conditions. I ran Prime95’s torture test with small FFTs for half an hour during each run. This was done a total of 12 times, increasing power limits at 10-watt intervals from 15 up to 35 watts. When I tried 45 W with all cases, the motherboard’s voltage regulator began throttling the CPU, or a thermal alert came up, so the 45 W tests were ended early. I changed CPU power limits quickly between each test so that the coolers lost as little heat as possible. This way I could get to heat saturation more quickly during each run. CPU core temperatures were tracked with HWINFO64. Over 17,000 core temperature readings were recorded and processed to compile these results. After a lot of averaging, I was left with this summary. The test system included an Intel i3-9100 CPU, Asus Prime H310T motherboard, 16 GB of Ripjaws DDR4 SODIMM memory from G.Skill, and a 500 GB 2.5-inch Crucial MX500 SSD. Let’s get to the thermal results before looking at a broader comparison of the four cases.
Below are the results for the Streacom FC2. The Y axis here is Celsius adjusted for a 25 Celsius room temperature. In other words, the ambient room temperature was subtracted from the average core temperature for each minute the tests were run, then 25 degrees was added back to mostly eliminate the room temperature’s effect. The FC2 passed the 35-watt test without any throttling, and began throttling during the 45-watt test. You can see here that the maximum temperatures increased by an average of about 16 degrees each time the power limits were increased by 10 watts.
These are the results for the Akasa Euler S. The Euler S again passed the 35-watt test, but began throttling during the 45-watt tests. Maximum temperatures increased an average of 15 degrees for every 10 watts of power use.
The Akasa Galileo T again passed the 35-watt test, but this time did not begin throttling during the 45-watt test. A voltage regulator thermal alert did show up though and the 45-watt test was ended early and is not shown here. Maximum temperatures with the Galileo T increased an average of 15 degrees for every 10 watts of power use.
Finally, here are the Turemetal UP2 results. Again, the case passed with 35 watts, but like the Galileo, a thermal alert came up and the 45-watt test was ended early. Maximum temperatures increased an average of 13 degrees for every 10 watts of power use. The maximum temperatures have gone down each time the case was changed, although the Euler S and Galileo T results are almost identical.
Below is a summary of results. Results between the Euler S, Galileo T, and UP2 are very close, but there was a clear increase in temperatures for the FC2. The Intel i3 CPU has thermal paste below its heat spreader or IHS, and a CPU that has a soldered heat spreader like Intel’s i7 or i9, or AMD’s higher end CPUs should have better results, although they might not make a lot of sense for this type of PC. Linear trendlines were added to the data, and formulas for stressed core temperatures were determined. If we use these formulas to calculate the theoretical maximum CPU power uses for each case, assuming an ambient room temperature of 25 degrees Celsius and maximum core temperatures of 100 degrees, we get 42 watts for the Streacom FC2, 48 watts for the Akasa Euler S, 49 watts for the Akasa Galileo T, and 52 watts for the Turemetal UP2. These results will of course vary somewhat with difference processors or different motherboards.
Something interesting happened between the two Akasa cases. With the Euler S, the VR began throttling after 4 minutes during the 45-watt test. With the Galileo T, I got a VR thermal alert, but no throttling after 4 minutes, even though CPU temperatures were about the same. I stopped the test after 8 minutes because of the thermal alert, so I don’t know if the VR would have throttled during the full 30-minute test. That may not sound like much of a difference, but to me this was significant. The Galileo case allowed the CPU to keep going longer at 45 watts. Even the Streacom FC2 did not cause VR throttling until 5 minutes into the 45-watt test. This to me indicated that the Euler is worse for motherboard component temperatures than the FC2, even though the FC2 was clearly worse for CPU temperatures. I think I know why. In the case of the Turemetal UP2, it has a more open design, with heatsinks in the front and on both sides. Air is allowed to easily enter through the corners of the case, which is great for motherboard mounted components. With the Akasa Galileo, it has a closed design, but, the heatsink is only to one side of the motherboard. Heat is transferred away from the motherboard, which is also good for motherboard mounted components. With the Akasa Euler, there is a large CPU block that makes contact with the CPU, and the heatsink wraps around the motherboard. There are vents on the sides, but it clearly is not enough to counteract the effect of having the heatsink that close to the motherboard. The Streacom FC2 also has a closed design, but it has a decent number of vents around it and it uses heatpipes to more quickly transfer heat away from the CPU.
Let’s look now at a broader comparison of these cases. The Turemetal UP2 seemed to do the best, and the Streacom FC2 the worst. The two Akasa cases were not far behind the Turemetal case though, although as discussed the Euler S was the worst for motherboard temperatures. The theoretical maximum powers and thermal resistances are listed here for each case, based on results with the i3 CPU used here. Although the Streacom case was the worst thermally, it is the smallest and lightest case, so it may be a good design for very low power applications. The two Akasa cases and the Turemetal case are similar in overall volume, although the Turemetal case is lighter than the Akasa cases. When considering installation difficulty, the Streacom FC2 and Akasa Euler were the easiest. The Galileo T is somewhat more difficult due a tricky heatsink disassembly and reassembly. The Turemetal UP2 is also somewhat more difficult due to a tricky heatpipe installation procedure. Unfortunately both the Streacom FC2 and Turemetal UP2 have been discontinued and are no longer available. I included them here because I thought their designs were still worth evaluating. The Akasa Euler S can be found on Newegg for $120, and the Akasa Galileo T can be found on Newegg for $123, including shipping to the United States. Other notes included here detail the case heatsink designs.
So, if you are considering a tiny fanless PC case for a low-power PC system, I hope this guide has been useful. Keep in mind that the CPU and motherboard chosen and room temperature will have a significant effect on CPU temperatures. An alternative to these Thin Mini-ITX cases include a somewhat larger Mini-ITX heatsink case such as those from Streacom, Akasa, Turemetal, or also from HDPlex. You can find a wide variety of custom built fanless PCs at fullysilentpcs.com.