![\"\"](\"https:\/\/www.hwcooling.net\/wp-content\/uploads\/2025\/03\/axagon-eem2-20g-01.jpg\")
<\/a><\/p>\nIncluded in the package \u2013 in addition to the metal body (aluminum), i.e., the enclosure itself \u2013 are a silicone sleeve, a thicker thermal pad (3 mm), a USB-C cable, and a screwdriver to remove the enclosure\u2019s cover. If you\u2019re concerned about having only one USB cable (with Type-C connectors on both ends), it\u2019s important to note that the supported 20-gigabit (USB) interface does not support any other connector type.<\/p>\n
While the enclosure will technically work with USB-A cables (on one end), if you acquire one, you\u2019ll need to reckon with a performance cap of 10 Gb\/s. That\u2019s the fastest Type-A interface used on motherboards. The USB 3.2 Gen 2\u00d72 standard supported by the EEM2-20G enclosure pertains exclusively to USB-C connectors. That\u2019s also why the package includes no cable other than the one with USB-C connectors.<\/p>\n For high performance, the enclosure\u2019s surface is perforated. This allows excess heat to dissipate through these openings. There are holes along the smaller walls and ventilation grooves along the longer sides of the enclosure. The largest surfaces \u2013 the front and back \u2013 are solid and sealed, with heat conducted into them via the thermal pad.<\/p>\n The sleeve avoids covering the enclosure\u2019s perforations (through cutouts) so it doesn\u2019t act as a thermal insulator. Its main function is to prevent mechanical damage to the enclosure due to accidental drops or greater external pressure. This sleeve can be removed from the enclosure. In fact, the product is shipped in this sleeve-free state.<\/p>\n The model name of the enclosure (EEM2-20G) is printed on the back, which is also the part where the most thermal exchange occurs. Heat is dissipated through the pad right here \u2013 at the rear.<\/p>\n After removing the back panel, the enclosure splits into two parts, with the main part housing the PCB with an M.2 connector.<\/p>\n Only SSDs with NVMe support are compatible \u2013 in four different lengths: 30 mm, 42 mm, 60 mm, and 80 mm.<\/p>\n To install a specific form factor, you\u2019ll need to correctly position a standoff post to secure one side of the SSD. On the other side, this function is handled by the M.2 connector into which the SSD\u2019s contacts are inserted. Once the SSD is mounted and before closing the enclosure, a thermal pad goes between the enclosure and the SSD \u2013 specifically between the cover and the SSD\u2019s controller and memory. When installing it, don\u2019t forget to remove the protective film from both sides of the pad.<\/p>\n A light guide for the indicator LED is missing from the SSD enclosure itself, but not from the PCB. The light shines through the perforation near the USB-C connector, making it easy to tell that everything is in order and that the enclosure is powered by the connected device.<\/p>\n The foundation for all measurements is CrystalDiskMark with a library size of 1 GB. We test SSD speed in two situations. Before load and then, after a 10-minute load for maximum sequential reads and writes. This will reveal how warming up the external enclosure affects the speed with the Samsung 980 Pro SSD (1 TB). The values of the speed measurements are averaged over three passes for greater accuracy. The enclosures are connected via a USB expansion card (3.2 gen. 2\u00d72) \u2013 the Renkforce RF-4538236 with the ASMedia ASM3242 controller. The tests run on an Asus ROG Strix Z790-E Gaming WiFi<\/a> motherboard with an Intel Core i9-13900K processor<\/a> (and G.Skill Trident Z5 Neo memory – 2\u00d716 GB, 6000 MHz\/CL30) under Microsoft Windows 10 (22H2).<\/p>\n We measure power consumption using a custom-made PCIe power meter, which you’ll also know from graphics card tests<\/a>. This is plugged in before the PCIe card (with a USB controller), whose consumption is part of the result. We measure the power consumption under load for maximum sequential read and write speeds.<\/p>\n We only observe the temperature on the surface, on the casing. Be careful when evaluating it, a higher temperature doesn’t automatically mean a worse result (think thermal pad tests and heatsink temperatures<\/a>…), but it doesn’t mean a better one either. The guiding factor here from a cooling perspective would be SSD temperature, but we have no control over that. The test SSD does have its sensors, but once it (the SSD) is plugged into the enclosure, the motherboard can no longer reach them. So we only do IR thermal mapping (with a Fluke Ti125 thermal imager), which can at least show how the heat is spreading through the enclosure and whether it will burn you. Thermal imaging is for the finned side of an enclosure or the side in contact with the SSD controller (if the enclosure does not have fins).<\/p>\n During testing, the SSD enclosures are always placed in the same position – longitudinally, in an open space, where they stand on the pad with the bottom side facing down. The airflow in the test room is always comparable, with the temperature at the control point varying between 21,0 and 21,3 \u00b0C.<\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a>\u00a0 \u00a0<\/a><\/p>\n<\/a><\/p>\nTesting methodology<\/h2>\n
<\/a><\/p>\n