E2 SSD Form Factor

Emerging workloads across AI, big data, IoT, and hyperscale compute require a rethinking of storage infrastructure. As data volume surges, particularly in the “warm data” tier, where information is accessed frequently but not in real-time, existing storage form factors struggle to scale simultaneously in terms of capacity, density, power, and performance.

Enter the E2 SSD form factor, developed jointly by the Storage Networking Industry Association (SNIA) and the Open Compute Project (OCP). Purpose-built for ultra-dense storage environments, E2 enables up to 1 petabyte (PB) of QLC NAND flash per device, leveraging PCIe 6.0 NVMe, within a compact thermal-optimized footprint.

Since the inception of flash storage, form factors have evolved from mechanical drive mimics (2.5” SATA) to purpose-built enterprise storage designs. These include M.2, U.2/U.3, EDSFF variants, and now the next frontier: E2.

The storage industry has experienced a remarkable evolution over the past five decades—from magnetic platters to NAND flash arrays—shaped by growing demands for capacity, reliability, speed, and energy efficiency.

In the 1980s and 1990s, storage was dominated by spinning disk hard drives (HDDs), prized for their low cost-per-gigabyte. As digital content, databases, and internet usage expanded, higher capacities were achieved through advances in perpendicular magnetic recording (PMR) and helium-filled enclosures, culminating in enterprise-grade HDDs reaching 20–60 TB in the early 2020s.

However, HDDs exhibited critical shortcomings:

• Latency typically in the millisecond range (5,000–8,000 µs)

• High failure rates in dense environments

• Significant cooling and vibration isolation requirements

This led to the rise of NAND flash storage in the early 2000s. With zero seek time, microsecond latency, and solid-state durability, SSDs transformed laptop, desktop, and enterprise performance. Early form factors like 2.5" SATA SSDs were drop-in replacements, constrained by SATA’s bandwidth (~6 Gbps).

The subsequent introduction of NVMe and PCIe-based SSDs in U.2, M.2, and later EDSFF formats unlocked:

• Direct CPU access via PCIe lanes

• Gigabyte-per-second throughput

• Sub-100µs latency

• Smaller thermal envelopes and better rack utilization

Despite these improvements, the growing volume of semi-active “warm data” exposed a market gap between high-cost performance SSDs and high-latency HDDs.

Key KPIs that shaped this transition include:

• IOPS per watt: SSDs dramatically outpace HDDs (e.g., 100K vs. 400 IOPS)

• Throughput per U: Increasing demand to store more per rack unit

• Cost-per-GB: QLC NAND and controller advances narrow SSD-HDD cost gap

• MTBF and reliability: SSDs exhibit fewer mechanical failures, especially in high-vibration environments

• Cooling efficiency: High-density SSDs consume less power per TB compared to HDDs at scale.

The introduction of the E2 form factor responds precisely to this intersection: the need to store massive datasets close to compute, with flash-grade reliability, at a cost that competes with HDD-based systems in warm-tier architectures.

As AI workloads, video surveillance, genomic data, and IoT sensors flood data centers with petabytes of mid-use data, E2 SSDs offer a scalable solution that fits within the thermal, mechanical, and economic envelope required by tomorrow’s data-driven infrastructure.