Solidigm D7-PS1010 1.92TB PCIe Gen5 SSD

The Solidigm D7-PS1010 1.92TB is a PCIe Gen5 x4 NVMe U.2 enterprise solid-state drive built on Solidigm's latest TLC NAND architecture, purpose-designed to meet the throughput and latency requirements of modern AI training pipelines, high-frequency analytics, and write-intensive storage tiers. Leveraging the doubled bandwidth ceiling of the PCIe 5.0 interface compared to Gen4, the D7-PS1010 delivers sequential read performance exceeding 14 GB/s and sequential write performance surpassing 7 GB/s, enabling data-hungry GPU clusters and inference engines to sustain peak utilization without storage bottlenecks. The drive's advanced controller integrates hardware-accelerated error correction, dynamic thermal management, and end-to-end data path protection to maintain consistent latency profiles under sustained queue depths.

The D7-PS1010 is positioned within Solidigm's datacenter-grade D7 product family, a lineage known for enterprise-class reliability and mixed-workload endurance ratings. At 1 Drive Write Per Day (DWPD) over its rated lifespan, the 1.92TB capacity point aligns with read-intensive to mixed-workload profiles such as AI dataset staging, checkpoint storage, and hot-tier object storage caching where both read throughput and write durability must coexist. The drive supports NVMe 2.0 command set features including Namespace management, SR-IOV-capable host memory buffer, and enhanced power state transitions critical for hyperscale and rack-dense deployments.

Launched in Q2 2025, the D7-PS1010 addresses the accelerating shift toward PCIe Gen5 storage infrastructure accompanying fifth-generation GPU and CPU platforms from major server ecosystems. Its 2.5-inch U.2 15mm form factor ensures broad compatibility with existing NVMe backplane designs while delivering next-generation bandwidth, making it a practical upgrade path for operators transitioning AI and HPC storage fabrics without full chassis replacement. The drive also incorporates Solidigm's proprietary firmware telemetry stack, enabling granular SMART attribute monitoring, predictive health analytics, and seamless integration with fleet management tools.

Ideal for

• AI and deep learning training dataset staging: providing sustained sequential read throughput to feed multi-GPU training nodes without I/O starvation during large-scale model training runs
• Checkpoint and model snapshot storage: absorbing high-bandwidth sequential write bursts during frequent gradient checkpoint operations in distributed training frameworks such as PyTorch and TensorFlow
• High-performance mixed-workload NVMe-oF target: serving as a backend storage device in NVMe over Fabrics disaggregated storage architectures requiring consistent sub-100µs read latency at scale
• Hot-tier caching in tiered object storage: accelerating metadata-heavy workloads and frequently accessed object retrieval in S3-compatible storage platforms deployed across hyperscale and enterprise private cloud environments
• Real-time analytics and OLAP acceleration: supporting columnar database engines and in-memory analytics offload scenarios where sequential scan throughput directly determines query completion time
• HPC scratch storage for simulation workloads: delivering the write endurance and bandwidth necessary for computational fluid dynamics, genomics pipelines, and seismic processing applications that generate and consume massive intermediate datasets

Technical specifications

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