Enterprise data centers are always trying to find the perfect balance between power, speed, and cooling within the limited space of a server rack. For storage, this challenge is even more critical. The issue often lies with traditional form factors, many of which were designed for older, slower hard disk drives (HDDs) rather than modern solid-state drives (SSDs). These older designs don’t provide the cooling or storage capacity that today’s data centers need.
The Evolution of SSD Form Factors
Initially, when enterprises began to adopt SSDs, manufacturers used familiar form factors like the 2.5-inch and 3.5-inch designs to make the transition easier. This allowed companies to upgrade their storage without having to overhaul their existing server racks. However, this compatibility came at a cost: the enclosures, originally built for HDDs, restricted airflow and limited the scalability of the faster, more efficient SSDs.
As technology advanced, SSDs began to appear in a variety of new shapes and sizes, including mSATA, U.2, and M.2. While these new form factors gave users more flexibility for general use, they still weren’t perfectly optimized for the high-demand, mission-critical workloads of large enterprise data centers. Organizations needed drives that offered superior capacity, performance, thermal management, and energy efficiency.
Introducing EDSFF: A Solution for the Enterprise
Around five years ago, the Storage Networking Industry Association (SNIA) developed the Enterprise and Datacenter Standard Form Factor (EDSFF) to overcome these limitations. The EDSFF family of drives is specifically designed to meet the unique needs of enterprise data centers. All EDSFF drives use the same edge connector, pinout, and functions, and they all utilize the NVMe protocol and PCIe interfaces for high-speed performance. The first two form factors introduced were the E1 and E3 series.
The E1.S, or “gum stick,” form factor is designed to replace M.2 drives in data centers. It offers higher density and power in a small size. One of its key benefits is that it’s hot-pluggable, allowing technicians to replace a drive without shutting down the system. A single 1U server can hold up to 32 E1.S drives, enabling data centers to easily scale their storage capacity.
The E1.L is a hot-pluggable, ruler-shaped form factor optimized for 1U servers. Its long design provides more board space for NAND flash packages, improving both capacity and cooling efficiency. Unlike older U.2 drives, which sometimes had folded PCBs that trapped heat, the E1.L spreads all components across a flat surface. This design requires 55% less airflow than U.2 drives and significantly reduces power costs. It’s also the densest storage form factor, with a single 1U server capable of holding a staggering 1 petabyte (PB) of data.
The E3 drive is designed as a modern replacement for the traditional 2.5-inch U.2 form factor. Although it has a similar shape, the E3 supports up to x16 PCIe lanes and 70W of power. It is hot-pluggable and optimized for both 1U and 2U servers, with four variations: E3.L and E3.S in both single (1T) and double (2T) widths. The thicker 2T variant is often used for high-power devices like computational storage that generate more heat.
The E3 is also remarkably flexible. It can accommodate a smaller E1.S printed circuit board (PCB) inside its enclosure, which helps reduce the cost for lower-capacity drives. This form factor also supports other device types besides SSDs, such as persistent memory.
Introduced in May 2025, the E2 form factor aims to bridge the gap between high-capacity HDDs and high-performance SSDs. It’s built for the “warm” data tier—information accessed too frequently for slow HDDs but not critical enough to justify the cost of top-tier SSDs.
The E2 is engineered for massive scale, supporting up to 1 PB per drive in a standard 2U server. A single chassis can house up to 40 drives, delivering a total of 40 PB of capacity while reducing space and power consumption. Its hot-pluggable and front-accessible design simplifies maintenance and minimizes downtime. While it has a higher power draw than HDDs, it offers significantly better power efficiency per terabyte, making it a more economical choice in the long run.
