SK Hynix PEB110 E1.S Enterprise SSD

The SK Hynix PEB110 is a PCIe 5.0 x4 NVMe E1.S enterprise SSD engineered to address the exponentially growing storage density demands of hyperscale data centers and cloud-native infrastructure. Built on SK Hynix's advanced 238-layer TLC NAND flash technology and paired with a purpose-designed enterprise NVMe controller, the PEB110 delivers sequential read throughputs exceeding 14 GB/s and sequential write throughputs surpassing 10 GB/s — nearly doubling the bandwidth ceiling of its PCIe Gen4 predecessors. The E1.S 9.5mm ruler form factor enables operators to populate far more drives per 1U chassis compared to traditional 2.5-inch U.2 configurations, dramatically increasing raw storage density without expanding rack footprint.

The PEB110 is optimized for mixed read/write enterprise workloads including AI/ML training data pipelines, large-scale object storage backends, distributed databases, and high-frequency transactional systems. Its end-to-end data path protection includes power loss protection (PLP) circuitry, advanced ECC with LDPC error correction, and hardware-accelerated encryption conforming to AES-256 and TCG Opal 2.0 standards. The drive is rated for 1 DWPD endurance across its full capacity range, making it suitable for both read-intensive and balanced mixed-use workloads in enterprise environments where data integrity and sustained throughput are non-negotiable.

Launched in Q4 2025, the PEB110 represents SK Hynix's commitment to the evolving EDSFF ecosystem, aligning with OCP (Open Compute Project) storage specifications and the growing hyperscaler preference for the E1.S form factor in next-generation server platforms. Its thermal design incorporates a precisely engineered aluminum heat spreader and supports active airflow management consistent with OCP-compliant chassis airflow standards, ensuring stable, throttle-free operation at sustained peak loads. The drive's low-latency random I/O performance — reaching up to 2.5M IOPS read — makes it equally adept at serving latency-sensitive NVMe-oF disaggregated storage architectures.

Ideal for

• Hyperscale object storage nodes requiring maximum NVMe SSD density per 1U chassis with PCIe Gen5 bandwidth headroom
• AI and machine learning training clusters demanding high-throughput sequential read performance for large dataset ingestion pipelines
• Distributed relational and NoSQL database deployments where sustained high random IOPS and low latency are critical for transaction processing
• Cloud-native Kubernetes persistent volume storage backends benefiting from NVMe-oF target deployments over high-speed fabrics
• High-frequency analytics and real-time data warehousing platforms that require consistent mixed read/write performance under 24/7 sustained load
• Virtualized server consolidation environments needing dense, power-efficient storage to support hundreds of concurrent VM storage workloads per rack unit

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.