Pavilion Data Systems

Pavilion Data Systems, established in 2014, carved a niche for itself in the high-performance data storage market before its acquisition by SK hynix in October 2022. The company was the brainchild of Kiran Malwankar, Sundar Kanthadai, and V.R. Satish, who brought together their extensive experience in storage and networking technologies. Malwankar, for instance, held significant engineering roles at companies like NetApp and OnStor, developing a deep understanding of the challenges in data-intensive environments. This collective expertise was foundational to Pavilion's mission.

The firm engineered a high-performance, dense, and scalable storage platform, explicitly designed from the ground up for NVMe and NVMe-over-Fabrics (NVMe-oF) protocols. This focus addressed the performance bottlenecks commonly found in traditional storage architectures, especially for applications demanding high throughput and low latency, such as those in AI/ML, HPC, and media and entertainment. Pavilion's core product, the HyperParallel Flash Array (HFA), was its signature offering, delivering performance comparable to or even faster than local NVMe storage but with the centralized management benefits of a Storage Area Network (SAN). The platform's architecture allowed for linear scaling, meaning performance increased predictably as more capacity was added.

Pavilion's business model revolved around selling these high-performance storage hardware and software solutions to enterprise clients, research institutions, and government agencies that manage massive, performance-sensitive datasets. Following the acquisition, the technology and engineering teams were integrated into SK hynix's U.S.-based storage solutions subsidiary, Solidigm. The goal of this strategic move was to incorporate Pavilion's advanced controller technology into SK hynix and Solidigm's existing portfolio of enterprise SSDs, creating more comprehensive and powerful storage solutions for data center clients. Keywords: data storage, NVMe-oF, high-performance computing, flash storage, enterprise storage, data center solutions, storage architecture, low latency, AI infrastructure, storage acquisition