Unigen Introduces Cheetah Enterprise SSD Series for AI and Hyperscale Infrastructure

Unigen Corp., a player in the design and manufacturing of enterprise and industrial electronics, announced the release of its new Cheetah series of Enterprise SSDs. Two complementary product families are now in production, designed for modern AI infrastructure:

• Cheetah Storage Class Memory (SCM) SSD: Ultra-low-latency SCM SSD designed to accelerate demanding AI and data center workloads
• Cheetah High Capacity (HC) SSD: Ultra-high-capacity NVMe SSD optimized for AI datasets and large-scale storage

Together, the solutions provide a unified storage architecture for AI servers, GPU clusters, hyperscale cloud deployments, and enterprise data centers, balancing high-speed, low-latency performance with dense storage capacity. Cheetah SCM: Speed for Critical Workloads Cheetah Storage Class Memory (SCM) are high-reliability, high-performance PCIe Gen5 x4 NVMe 2.0 SSDs designed for next-gen enterprise and data center storage. Rated for up to 60 DWPD for random workloads and 80 DWPD for sequential workloads, Cheetah SCM delivers the endurance needed for sustained, high-intensity data operations. With 16 NAND channels and a DDR5 interface, it also delivers ultra-low latency and high random IO/s for demanding AI and data center workloads. This combination of endurance and performance makes Cheetah SCM a differentiated storage solution in the QLC era of high-capacity storage.

Cheetah SCM bridges the performance and persistence gap between memory and traditional NAND flash, providing higher storage capacity than DRAM while maintaining fast, consistent access to critical data. Available in U.2, E1.L, E1.S, and E3.S form factors, Cheetah SCM provides a flexible, high-integrity storage platform for next-gen enterprise and data center architectures.Cheetah HC: High-Density Storage for Scale Cheetah High Capacity (HC) are high-performance, high-capacity PCIe Gen5 x4 NVMe 2.0 SSDs designed for next-gen enterprise and data center storage. With 16 NAND channels, a DDR5 interface, and advanced security features, Cheetah HC delivers high throughput, low latency, and enterprise-grade reliability.

Supporting capacities up to 122 TB in U.2 and E1.L form factors, Cheetah HC is optimized for AI workloads, data lakes, object storage, and other applications requiring high-capacity, high-density storage. Better Together: The Ideal AI Storage Hierarchy Deploying Cheetah SCM and Cheetah HC together creates a two-tier storage hierarchy designed to optimize throughput and manage overall storage costs. Cheetah SCM handles active datasets, write acceleration, caching, and temporary scratch space, while Cheetah HC stores foundation models, training datasets, embedding databases, and long-term archives. Purpose-Built for AI Infrastructure The Cheetah series is designed for:

• AI training clusters
• LLMs
• GPU inference farms
• Hyperscale cloud storage
• HPC
• Enterprise storage arrays
• Vector databases
• Disaggregated storage architectures
• GPU Direct Storage deployments

Availability Both Cheetah SCM and Cheetah HC are in production.