Are M2 SSDs Becoming Server-Ready?

By WPG AmericasNovember 28, 2016

The M.2 SSD is a solid-state drive (SSD) designed to deliver high-performance storage solutions in space-constrained environments such as the ultrabook and tablet computers, as well as data center servers. Developed from the ground up, M.2 supersedes the considerably larger mSATA form factor, enabling a leading-edge in packaging density with double-sided configurations as compared to its single-sided predecessor.

M.2, formerly named by the PCI-SIG consortium of technology as the Next Generation Form Factor (NGFF), supports critical applications like Wi-Fi, PCI Express (PCIe), Universal Serial Bus (USB), and Serial ATA (SATA).

M.2's Promising Capacities Might Just Make it Server-ready

The recent developments in Open Compute Project (OCP) are bolstering M.2's increasing proliferation into hyper-scale data centers. The OCP movement is dramatically redefining data center operations, prompting more and more hyper-scale switch overs to M.2 for optimized storage solutions.

The capabilities of M.2 NVM Express (NVMe) drives are being upheld by a number of positive announcements that supplement the SATA, which has been in the market for some time now.

An M.2 connector facilitates four lanes of PCI Express (PCIe) 3.0 bandwidth, or alternatively, either a single SATA or a USB 3.0 connection. Support for the latest PCIe 4.0 is in the pipeline and might be released as we talk. This implies an abundance of bandwidths to keep up with the fastest of SSDs.

Hyperscalers Are Deploying M.2 SSDs for Advanced Competitive Edge

The growing adoption of M.2 devices lately featured on the ‘In Tech We Trust’ podcast, The Flexible Cloud, which highlighted how big hyperscalers have picked up on M.2 SSDs to cram high-efficiency storage into a single server.

Tech giants, the likes of Facebook, Google, and Microsoft Azure are using M.2. This, by itself, serves as an assurance for cost-effectiveness, as these data sensitive enterprises are extremely cost-conscious from both operational as well as infrastructural perspectives.

Upcoming data centers and also the existing ones can completely rely on the cost-profit analysis of these companies, knowing that they have thoroughly reviewed the cost point before migrating to M.2s. Presentations from SNIA Developer’s conference and the Flash Memory Summit covering M.2 further reinforce similar stories.

To Go With One Mega-SSD or Multiple Smaller Devices?

M.2, with releases lined up for next-generation storage capacities of 60 to 100TB, portrays the diversity we now notice in the NAND flash market.

Though there's no direct accumulation in cost saving, a larger drive can, however, be a lot easier to service than numerous smaller ones. M.2s basically pay off in the long run, considerably on the maintenance part.

Since costs are primarily focused on the number of NAND chips in an SSD, a linear reduction in cost can only come from increased density in each NAND chip.

The highest capacity of an M.2 currently available off-the-shelf is 1TB, but announcements of 100 TB 2.5-inch M.2 SSDs at the Flash Memory Summit clearly reflect the future of hyper-converged systems that integrate exceptionally optimized storage capacities in drive sizes similar to a stick of gum.

To know more about this product, please contact the WPGA team.