decorative mesh fabric gradient

Performance and Density at Scale

The unique architecture of the Pavilion HyperParallel Data Platform gives customers unprecedented choice and control to easily deploy high performance, low latency storage with unlimited scale for workloads across block, file, and object workloads without compromises.

Multifabric

Unrivaled Performance Density

Achieve business outcomes that were previously not possible with industry leading performance, throughput, and ultra low latency. Simplify your data center and reduce costs with the only solution that delivers industry leading performance for block, file, and object workloads in a compact 4RU system.

Modular

Performance and Availability

Never choose between performance and availability again. The unique Pavilion architecture enables the use of active standby controllers, so customers get the performance of multiple controllers and the security of redundancy.

Lightening

Performant Unified Storage

Get the right performance for each workload with up to 20 controllers per system, each capable of natively running block, file, or object workloads for maximum performance.

Check

Unlimited, Linear Scale

With up to 60GB/s read and 40GB/s write performance for NFS and S3 workloads, across an unlimited number of systems, customers can scale capacity or performance linearly and independently.

Hardware Top Down
SSDs
Line Cards
Dual Mgmt Controllers
Redundant Fabric
Modular Cooling

Our Hardware

Built with standards based components and no custom ASICs or FPGAs, customers enjoy the benefits of affordability without the headaches and compatibility problems of custom built hardware.

Standard NVMe SSDs

With up to 72 2.5” NVMe SSDs, customers can use TLC and SCM based NAND to get the best combination of performance, endurance, and affordability with up to 2.2PB of capacity per array.

Active-Active Line Cards

Up to 10 line cards, each with 2 independent controllers, each with its own CPU, memory, network, and OS instance, with either 4 100Gb Ethernet or Infiniband ports or 1 200Gb port.

Dual Management Controllers

A sophisticated GUI, CLI, and APIs for integration with V-Center, Kubernetes CSI, RedFishTM and SwordFishTM.

Redundant Fabric Modules

Dual, hot-swappable 1.6 Terabit fabrics form the core of a storage array designed like a network switch.

Modular Power & Cooling

Redundant, hot-swappable power and cooling modules keep you powered up and cool.

Hardware Top Down

Our Hardware

Built with standards based components and no custom ASICs or FPGAs, customers enjoy the benefits of affordability without the headaches and compatibility problems of custom built hardware.

SSDs

Standard NVMe SSDs

With up to 72 2.5” NVMe SSDs, customers can use TLC and SCM based NAND to get the best combination of performance, endurance, and affordability with up to 2.2PB of capacity per array.

Line Cards

Active-Active Line Cards

Up to 10 line cards, each with 2 independent controllers, each with its own CPU, memory, network, and OS instance, with either 4 100Gb Ethernet or Infiniband ports or 1 200Gb port.

Dual Mgmt Controllers

Dual Management Controllers

A sophisticated GUI, CLI, and APIs for integration with V-Center, Kubernetes CSI, RedFishTM and SwordFishTM.

Redundant Fabric

Redundant Fabric Modules

Dual, hot-swappable 1.6 Terabit fabrics form the core of a storage array designed like a network switch.

Modular Cooling

Modular Power & Cooling

Redundant, hot-swappable power and cooling modules keep you powered up and cool.

“Cisco’s ACI fabric with zero trust security and advanced traffic congestion control methods is a perfect fit for NVMe storage over ethernet. At Cisco’s NVMe research and testing lab, we have been using Pavilion HyperParallel Flash Arrays to test tens of millions of storage IOPS over NVMe-TCP sessions with Cisco UCS servers. The ability to change protocols with the click of a button delivers unrivaled ease of use. The Pavilion HyperParallel Data Platform with multi-controller support, flexibility, ease of use and management, sets the industry performance standard for NVMe data storage,”

Kamal Bakshi,

Director, Cisco Systems

Discover what’s possible. It’s easy to get started

Contact Us  –  (669-263-6900) – info@pavilion.io

1. Define Objectives

Tell us what you want to achieve and we will set up a call with a Systems Engineer to determine requirements.

2. Test the System

Remote and on-prem POCs to demonstrate unmatched results for Block, File, and Object at any scale.

3. Deploy Pavilion

Shatter expectations by delivering unmatched performance and density with ultra low latency.