
PDF version of this report
You must have Adobe Acrobat reader to view, save, or print PDF files. The reader
is available for free
download.
The M.2 Connector
Copyright 2021, Faulkner
Information Services. All Rights Reserved.
Docid: 00021031
Publication Date: 2107
Report Type: TUTORIAL
Preview
With the trend in computing moving continually towards smaller, lighter, and
more integrated systems, new technologies are being sought that can help
a PC do more while taking up less space. To this end, the Serial ATA
International Organization, which governs the SATA standard, and the Peripheral
Component Interconnect Special Interest Group (PCI-SIG) joined efforts to
create a more compact, faster, and more versatile connector standard for PC
motherboards. This came at a time when ultrabooks were largely replacing older,
larger laptops while still being required to meet or exceed the speeds of
their predecessors with just a fraction of the space to work with. To fulfill
this need, something even more diminutive and adaptable was necessary to
connect storage, networking, and other devices to the slim and light ultrabooks
being produced. This is where the M.2 standard (formerly known as Next
Generation Form Factor) comes in. Not only is it capable
of doing all of the same jobs as the older, more well-known mini-SATA (mSATA)
and PCI Express Mini Card connectors, but it is able to do so while providing
higher maximum data transfer rates and throughput capacities than either of its
predecessors. These characteristics, combined with a form factor that barely
takes up more room than a soldered-on component, has resulted in the rapid and
widespread adoption of M.2 not only in ultrabooks but also in a growing number
of compact desktop PCs and even larger tower systems. This report will examine
the parameters of the M.2 connector, how it can be used, the versions of it
currently being installed in systems, and its technical specifications.
Report Contents:
Executive Summary
[return to top of this report]
| Related Faulkner Reports |
|
Internal Flash Storage and Solid State Technology Wireless Local Area Network |
The M.2 standard was born when something smaller and more versatile than
either mSATA or the PCI Express Mini Card connector was needed in compact computers
– particularly in thin-and-light ultrabooks and mid-range laptops. This connector had to meet or exceed the speeds offered
by its predecessors while taking up less space and providing a wider
range of functionality than both previous types combined. To handle these lofty
expectations, M.2 was designed to serve as an onboard interface for not
just storage devices but for various forms of networking peripherals as well.
While the devices M.2 is capable of connecting could be soldered directly
onto a motherboard or mainboard, this would make it nearly impossible to upgrade
them at a later date. Since onboard storage is one of the most commonly
upgraded portions of a PC’s internal components, this would prove less than
ideal and would likely sit poorly with the portions of the consumer market are
unwilling to entirely surrender their right to upgrade the machines they own.
That said, while
the average customer cannot afford to simply discard their whole PC because
they needed some extra storage space, some companies, like Apple, have forgone
the M.2 connector in favor of soldered components. The vast majority, however,
have not been successful at convincing consumers that they should have no
control over upgrades. Due to its ability to continue providing upgradeability
in thin and light laptops, M.2 became popular
very quickly, as it could provide the necessary data and power connections
to a peripheral while taking up almost no more room than a few solder joints.
Additionally, it provided an interface that is no more difficult to swap components
into than plugging in a flash drive. However, M.2's specifications were
attractive enough that its adoption quickly spread beyond the thin-and-light ultrabooks for which it was intended and into a much wider range of products,
including compact desktop PCs and even some larger tower systems. The last of
these installations may be the most surprising given the room available to such
large systems. But, to understand why, one need only realize that the two governing
organizations above completely overshot their goals in designing M.2 when it
came to the ability to handle high-speed data transmission. M.2
could provide faster throughput than not only mSATA and PCI Express Mini but
also faster than the 6.0 gigabit per second (Gbps) SATA interface that had been
connecting most desktop storage solutions for many years by this point. The new
option for a storage connector could not come at a better time, as the
transmission speeds offered by many top-of-the-line solid state drives (SSDs)
had already outstripped the 6Gbps standard and was being bottlenecked by its
constraints.
Description
[return to top of this report]
Figure 1. A close-up of an M.2 connector
Source: Newark.com
Applications
To understand the importance of M.2 technology,
one must first understand the variety of uses made possible by the
connector. The fact that there are a variety of these use cases is what makes
the M.2 connector so special. Not only can it serve as an interface for storage
devices (typically SSDs), but it is also able to connect and power nearly
any peripheral computing device thrown at it, including networking,
Bluetooth, display and audio adapters, as well as lighting controllers, and
on, and on. The reason for this flexibility is M.2's ability to replace
three of the most commonly used interface connector types found on most
motherboards: PCI Express, Serial ATA, and USB. M.2 is able to connect and power
any compatible device that would previously have used any of these three
connectors. This is not to say that all three of those devices types can simply
be plugged into any M.2 connector. Rather, each device needs to have an
appropriate male connector itself in order to slot into the M.2's port. This
makes M.2 the only type of connector that can support all three of the
aforementioned protocols interchangeably, allowing the same M.2 port on a
given motherboard to, over the life of that motherboard, be used to connect a SSD, a display adapter, a Wi-Fi card, and an audio adapter.
Even with this level of versatility, only
software-side changes would need to be made between each alteration. This example
should begin to reveal the incredible potential offered by this connector
technology.
Below is a more in-depth look at each the three
interfaces supported by M.2, including their parameters and what types of
devices they are typically used to connect.
-
PCI Express –
Peripheral Component Interconnect Express, more typically written as PCIe,
is a modern interface standard that replaced older PCI and Accelerated
Graphics Port (AGP) protocols. As the name of one of its predecessors would
suggest, it was typically used to connect devices designed for graphical
applications. Specifically, it connects and powers most display adapters
found in modern desktop PCs. As with most PC interfaces, it has gone through
various iterations, the most common version of which is still PCIe 3.0. PCIe
4.0-enabled devices are finally starting to come to market after several
years of waiting, but they still have nowhere near the market penetration of
PCIe 3.0, which is also the version that is most commonly replicated
by an M.2 connector. The 3.0 version is capable of data transfer rates of up to 985
megabytes per second, per lane.1 The "lane" refers to a specific
data pipe on the motherboard, multiples of which can be assigned to each
connector with a maximum of 16 lanes available to a given PCIe 3.0
connector. By multiplying the available lanes, it is possible to provide a
total maximum throughput of 15.75 gigabytes per second to a single PCI 3.0,
or, in this case, a M.2 connector.2 The massive amount of
data that can pass through an interface of this type has made it ideal for
powering even the most high-end display adapters, while also making it
possible to provide enough throughput for a storage solution connector as
well. This is why several companies (Intel, Corsair, Kingston, Asus, and more)
created SSDs that plugged into PCIe ports typically reserved for display
adapters. These units did provide improved throughput when compared to
legacy SATA solutions for high-end gaming PCs, video editing workstations,
and other demanding jobs. However, they took up even more space
within a system than a typical 2.5-inch drive would have, posing issues such
as reduced airflow and increased space requirements. This rendered such
options effectively useless for laptops and ultrabooks, leaving them out
in the cold until M.2 came along to fill the same role at a fraction of the
size. The protocol's compact size makes it theoretically possible for
motherboard manufacturers to add an M.2 port to their systems with the
capacity to provide the full 15.75 GB/s mentioned above, all while fitting
within the diminutive confines of something like an ultrabook or compact
desktop. Although current implementations max out at 4x (or four lanes for a
total speed of about 3.9GB/s), new protocols are in development that will
remove this limitation. Looking forward, M.2 technology can also support the
aforementioned PCIe 4.0 standard. Although the technology is just now
trickling onto the market, it has already been seen in a few ultra-high-end
motherboards, with M.2 support coming along for the ride. This
means M.2’s throughput potential will essentially double, once the new
protocol reaches widespread use in motherboards and components. It should also be noted that, while it is
technically possible to use the M.2 port to connect a display adapter to a
compact PC or ultrabook, the size requirements and cooling requirements of
most modern display adapter solutions would make their placement in such a
device impractical. That said, it would be entirely possible for such a use
case to occur within a larger tower system, but it remains a theoretical
scenario due to the fact that such systems include ample room to support
multiple large PCIe-connected devices. -
Serial ATA 3.0 – This protocol is the one
that M.2 is perhaps most commonly used to replace. Thankfully for legacy
applications, it is fully supported by M.2 connectors. However, its
availability on the newer connector type does not remove the 6.0 Gbps
limitation placed on the technology by its own protocol. That said, it is
possible for M.2 interfaces to support certain older SSDs and other flash
storage solutions which were designed for SATA 3.0, typically via additional
adapters. While this portion of the M.2 connector's versatility does not
really extend its usability into the future, it does expand it into the
past, making it a viable option for system builders that are required to use
legacy hardware within their systems. -
USB 3.0 – This use case, more than the
other two combined, provides M.2 technology with great versatility.
Just as an external USB 3.0 connector can add a nearly limitless range of
peripherals and expansions to a PC, so too can an M.2 connector expand a
system. While the transfer rate is even slower than SATA 3.0 at 5 Gb/s,
this is more than enough for most of the uses to which USB 3.0-connected
peripherals can expect to be put.3 The most common of these
USB-powered uses include Wi-Fi adapters, Bluetooth adapters, and cellular
modems. While all of these solutions can be permanently attached
to a motherboard, particularly within a laptop or ultrabook, the option to
upgrade at a later date by simply plugging in a new card to an M.2
port is a huge boon to both consumer-level and enterprise-class PC shoppers.
For instance, even with the constantly evolving technologies behind wireless
networking, an entirely new type of wireless LAN could potentially be
added to an older system via an M.2 port where a laptop with a soldered-on
Wi-Fi card would be left out in the cold.
Versions
While M.2 as a whole is nearly limitless in its potential, the
current versions of it are more task-oriented, with each version having a
specific range of size requirements and interfaces available
to it. These various editions are divided into alphabetically assigned "keys"
that determine the aforementioned parameters, with each of the currently used
keys powering a particular set of common use cases. That said, the
number of keys available is still expanding with the potential for more
iterations of the M.2 protocol coming in the future. This insures that, even as
computing requirements become even more stringent in the future, M.2 will be
able to adapt to growing needs. Returning to the present, it is useful to look
at each of the currently available common key types, including their size
parameters, supported interfaces, and common uses.
It should be noted that the measurement parameters for each
key type are defined in a very specific way via a four or five digit code. This
code can be read in millimeters, with the first two digits representing the
width of the card while the remaining two or three digits represent the card's
length. For example, an "A key" M.2 module can have a dimension code of 2230,
meaning that it is 22mm wide by 30mm long. The dimensions of the card are
extremely important not only for the amount of room which each unit takes up,
but also for the support screw that has become a standard part of all current
M.2 connected devices. This screw is similar to a motherboard standoff screw in
that is designed to provide a gap between the motherboard and the connected
device while also securing it in place to remove the potential for becoming
disconnected when its system is physically moved. Thankfully, many motherboards
with an M.2 port include several threaded holes into which the necessary
standoff screws can be installed, allowing for several key types to be supported by a
single board. The image below shows an example of these optional holes, with
each marked by a number representing the length of the card it should be used
for.
Figure 2. An M.2 Connector (Top Right) with its standoff screw installed.
Source: LinusTechTips.com
Key Types4,5
In addition to the size of the card, there is another
compatibility concern that must be considered, which is the pin configuration.
For example, A key and E key M.2 cards have an identical range of supported
physical sizes but different pin configurations. While this may seem
confusing, it has thankfully been made relatively easy due to the fact that most
manufacturers stick with a single type of connector based on the types of M.2
cards which are most likely to be connected to it. While this does limit the
range of usability, it also makes it easier for shoppers to be
confident that their devices will be compatible.
Below is the full list of currently defined key types, about
half of which remain unused as they are reserved for future applications. This
is noted where appropriate.
A
- Measurements – 1630, 2230, 3030.
- Available Interface Types – PCIe
x2, USB 2.0, DisplayPort x4. - Most Common Uses – Wireless Networking
Cards, Cellular Modems.
B
- Measurements – 3042, 2230, 2242, 2260,
2280, 22110. - Available Interface Types – PCIe
x2, USB 2.0 and 3.0, and Audio protocols such as PCM (Pulse Code
Modulation). - Most Common Uses – SSDs and Audio
components.
C and D
-
Reserved for future applications.
E
- Measurements –
1630, 2230, 3030 - Available Interface Types – PCIe
x2, USB 2.0, and a range of limited-use technical connectors, including
SDIO, UART, PCM - Most Common Uses – Wireless Networking
Cards, Cellular Modems.
F
-
Reserved for future applications including the "Future Memory Interface (FMI)"
G
-
Unused
H through L
-
Reserved for future applications.
M
- Measurements –
2242, 2260, 2280, 22110 - Available Interface Types – PCIe x4,
SATA, SMBus - Most Common Uses –
SSDs
Maximum Component Thicknesses6
The final consideration for M.2 card parameters is the card
thickness. This includes single-sided cards, where components are mounted on
only one surface of the card, as well as double-sided cards with components
on both sides. These are not dependent on the key type, with varying
thicknesses manufactured within varying keys. Specific size notations are show
below with the "S" representing single-sided modules, while the "D" represents
double-sided units.
Single Sided
- S1 – 1.2mm thick on top side.
- S2 – 1.35mm thick on top side.
- S3 – 1.5mm thick on top side.
Double Sided
- D1 – 1.2mm thick on top side, 1.35mm thick
on bottom side. - D2 – 1.35mm thick on top side, 1.35mm
thick on bottom side. - D3 – 1.5mm thick on top side, 1.35mm thick
on bottom side. - D4 – 1.5mm thick on top side, 0.7mm
thick on bottom side. - D5 – 1.5mm thick on top side, 1.5mm
thick on bottom side.
Current View
[return to top of this report]
As the market stands currently, M.2 technology and connectors remain a
protocol that is generally seen in high-end systems. This fact is true for ultrabooks
(which tend to be on the higher end of the laptop market) and compact desktops, and even more true for larger tower PC systems
where the benefits of M.2 technology are more of a luxury than a necessity.
Because of this reality, the number of products being produced that support the
protocol are still smaller than those using the more traditional 3.5-inch HDDs
and SSDs that use more well-established connectors, like SATA. The market is, however,
growing, with most major storage manufacturers having already gotten on board.
Storage companies currently offering M.2-based SSDs include Samsung, Crucial,
SanDisk, Kingston, Intel, AData, and more. As with any newer technology, pricing
currently displays a relative premium over comparable units with older
connectors, particularly in the case of M.2 SSDs versus SATA equipped models.
That said, this extra cost does come at the benefit of additional speed and
reduced space requirements. Whether or not this additional cost is justified is
something that each system build or upgrader will have to decide for themselves,
particularly if they have the room for a larger, typically cheaper module based
on the PCIe platform, which can provide the same data throughput speeds as M.2.
Outlook
[return to top
of this report]
For many years now, the PC market has been moving towards using laptop or
ultrabook systems as the primary computer for both private and business users.
This is because of their additional versatility, portability, and simplicity.
For this reason, M.2 and its aforementioned benefits for systems with limited
internal dimensions, will likely become even more important in the future. While
PCIe can provide essentially the same benefits for a full-tower desktop PC,
these types of systems are becoming less and less common, with most new models
being geared almost exclusively towards power users of one type or another, such
as PC gamers, video editors, 3D modelers, or some other profession or hobby that
requires ample processing, storage, and speed. For these power users M.2
does provide the benefit of simplicity, and increased airflow in their PC case,
but it is not the vital advancement for the desktop market that it is for the
majority of users populating the laptop and compact PC market. That said, future
full-tower system builders may still wish to go with a M.2 drive for its
simplicity of installation, as well as the aforementioned ease of air flow.
After all, systems that are powerful enough to warrant the inclusion of an M.2
drive tend to be hot enough to warrant taking airflow into consideration.
It is still relatively early days for M.2 technology. Like any new protocol or connector
this early in its history, it still has the potential to succeed or fail based
on what comes along to replace it, as well as which motherboard and component
manufacturers choose to adopt it. That said, any replacement for the technology
would need to offer impressive benefits at this point in order for manufacturers
to abandon M.2 technology. PC makers such as Dell, Acer, Asus, Gigabyte, MSI,
and many other high-end laptop makers have come to use M.2 drives as the primary
form of solid state storage shipping with their premium laptops. Even larger,
desktop replacement units now often sport M.2 drives thanks to their reduced
thermal impact and space saving characteristics. Although it still is not quite
as well established as its older counterparts mentioned above, the future of M.2
technology looks poised to continue growing for the next several years, at
least.
References
- 1
"PCI Express 3.0 Bandwidth: 8.0 Gigatransfers/s." ExtremeTech. August 2007.
2 Ibid.- 3 "USB 3.1 Specification Language Usage Guidelines." USB.org, March 2016.
- 4 Andrew Cunningham.
"Understanding M.2, the Interface That Will Speed Up Your Next SSD."
Ars Technica. February 2015. - 5 "M.2 Connector (NGFF) Introduction." Attend. February 2014.
- 6 Ibid.
Web Links
[return to top of this report]
- PCI-SIG: http://www.pci-sig.com
- The Serial ATA Organization: http://www.sata-io.org
About the Author
[return to top of this report]
Michael Gariffo is an editor for Faulkner Information Services. He
tracks and writes about enterprise software, the Web, and the IT services
sector, as well as telecommunications and data networking.
[return to top of this report]