For the past several years, mSATA has been the standard for people looking to add compact, solid state storage to their hardware. In the same vein, mini PCIe has allowed easy expansion in smaller computers and mini ITX motherboards. However, a new standard is emerging that could dramatically improve how we use internal expansion. This new standard, called M.2 (formally Next Generation Form Factor, or NGFF), is featured in our new ML100G-30.

But what is M.2, what is is used for and how does it work?


The first important thing to note is that M.2 is physically different than Mini PCIe, meaning that older mSATA drives and expansion (like wireless cards) will not work in M.2. It is unique because the standard has a wide variety of sizes for its cards. Also, it has several different kinds of keys. Keys are notches in the card to prevent the wrong kind of device from being plugged in. With several different lengths and many keys, it can come in dozens of different variations. At first glance, this may seem like a superfluous and unnecessary change,however M.2 is so robust from a technological standpoint that it requires these multiple variations.


The technical reason behind the keys rests in the fact that M.2 can take advantage of the PCIe, USB, Display, Audio, I2C or SATA busses on the motherboard. The specific motherboard wiring to the M.2 slot determines what kind of key it gets (a single card  can use multiple keys), preventing an incompatible card from being plugged into a motherboard. For example, their are 4 common types of keys being used today and each is assigned a letter for identification. The table below shows how each of those keys varies:

M.2 Keying Chart


So what does all of this keying and bus interoperability mean for the average computer? In a nutshell it will lead to faster, more flexible and more available storage and expansion. As an example, take SSDs. Before M.2, SSDs worked off of the SATA bus. The latest SATA technology can transfer data at a max of 6Gb/s (Six gigabits per second). This is pretty fast, but not even close to PCIe. PCIe can transfer data at 8Gb/s per lane (the number after the “x” in PCIe is the number of lanes, so PCIe x2 has two lanes). This means that a PCIe based M.2 SSD drive can transfer data at up to 5x the speed of a traditional SSD. These same improvements are true for other expansion devices as well.

M.2 Speed Graph


Currently the biggest issue with M.2 is the general lack of support. With that being said, M.2 is fairly new, and it’s not as if there are signs that it won’t catch on. In fact, this is the standard for all new Intel NUCs, and many mainstream motherboards are starting to include it. As adoption rates continue to climb, we expect to see many new devices starting to crop up. Once that happens, look for it to become a big player in the Industrial and Embedded computing sector.

One other issue M.2 will have is interoperability. As we see adoption grow, we may see different motherboard manufacturers using different keying. This may lead to devices that are specific to only a few manufacturers, or worse, it could cause stalled adoption due to a fractured ecosystem. This however is unlikely as Intel is usually very good at encouraging adoption of new standards and acts as a steadying force for the market.


Personally, I am really excited about M.2. Significantly more performance combined with increased choices of expansion that are easier for manufacturers to adopt? That sounds like a recipe for a winning technology standard.

To see M.2 in action, make sure to check out the brand new ML100G-30 Ultra Compact Fanless NUC featuring M.2 storage and wireless.