Vexata VX-100F Scalable NVMe Flash Array
NVMe-based Scalable Data Infrastructure for the Modern Enterprise
Contact us for pricing!
As enterprises transform to cloud and mobile business models, database platforms are being pushed to process higher volumes of complex events in real time, add users, and simultaneously run predictive analytics and machine learning on very large datasets.
This puts tremendous pressure on legacy storage arrays to handle larger data volumes without compromising IOPS, throughput, and I/O latency. As enterprises deployed allflash arrays, they experienced significant performance degradation due to the legacy storage controller architectures.
Enter the Vexata VX-100F Scalable NVMe Flash Array, which delivers over ten times the performance at lower costs than conventional all-flash storage arrays. The VX-100F enables the enterprise to experience massive improvements in application efficiencies and provides the blazing performance needed for machine learning/AI, next generation analytics, and other transactional applications that are used in the transformation to the cognitive era.
The VX-100F supports Fibre Channel attached Non-Volatile Memory Express (NVMe,) giving IT the flexibility required to cost-effectively deploy across a broad number of transactional, AI and analytic use cases.
The V X-100F uses a parallel, high availability architecture with active/active controllers that fully utilize the NVMe flash media. VX-100F delivers outstanding I/O performance, ultra-high throughput with ultra-low latency, 99.9999% availability, enterprise data services, and world-class support at an affordable price point.
The V X-100F supports 3 to 16 hotswappable storage blades, called Enterprise Storage Modules (ESMs). Each ESM contains four 3.84TB or 7.68TB NVMe SSDs, supporting expansion to 491TB of raw capacity in a 6U chassis.
The VX-100F is powered by the Vexata Operating System (VX-OS), delivering advanced data services including thin provisioning, pattern removal, always-on encryption, and space-efficient snapshots and clones. All data at rest is encrypted with Advanced Encryption Standard (AES) 256-bit encryption, which operates without any performance degradation.
The VX-100F uses RAID 5 and RAID 6 data protection, with unparalleled rebuild times on failure compared to conventual all-flash arrays. The VX-100F provides enterpriseclass resiliency, high availability, and non-disruptive upgrades (NDU) of software and hardware.
Even under impaired conditions, the VX-100F ensures that access to data is available 99.9999% of the time, while ensuring that the overall performance impact is minimal and there is no disruption to application I/O.
Enterprise Use Cases
- High Transaction Databases
- Business Intelligence
- Big Data & Real-Time Analytics
- Machine Learning/AI
- Time-Series Databases
- Active-Active Controller HA
- RAID 5/RAID 6 Protection
- Continuous Operations with 99.9999% Uptime
- Non-Disruptive Upgrades
- FC Block Interfaces
- Web UI, CLI, RESTful API
- Call Home with Proactive Support
- Thin Provisioning
- Pattern Removal
- Space Efficient Snaps/Clones
- Data at Rest Encryption
Rethink your data infrastructure
Today’s enterprise relies on data and business intelligence to create competitive advantage, but must be able to accelerate speed and scale of digital transformation to maintain that edge.
High Performance Access for Unstructured Data and Distributed File Systems
When deploying high performance solutions to support use cases for unstructured and file system data, you need a system that provides real-time access to unstructured data sets. The Vexata VX-100FS Scalable NVMe File Array is a tightly integrated and engineered Scale-out File Controller integrated with a VX-100F Scalable NVMe Flash Array. The VX-100FS provides unparalleled NFS, GPFS and HDFS performance, enterprise class resiliency, rich data services and scalability to uniquely solve the emerging Video, EDA, Genomics, IoT, Machine learning and AI challenges.
Sample Use Cases
Financial Trading – Time is money. Vexata delivers enterprise grade storage for a broad range of high performance applications used widely in financial trading. By removing latency from the I/O stack, trades happen faster with dramatically better analytics accuracy.
Risk & Fraud Analytics – Businesses use machine learning to analyze and interpret patterns and structures of data to reduce risk and increase their fraud-detection efforts. Vexata accelerates these workflows.
Artificial Intelligence / Machine Learning – Businesses use AI/ML powered by massively parallel GPU servers such as the NVIDIA DGX-1 to accelerate training and inferencing time, but these systems are impacted by storage performance. VelocityAI is ideal for accelerating AI/ML/DL workloads.
IOT/Analytics – Data is doubling every two years, driven by human and machine generated data, such as the Internet of Things (IoT). The analysis of patterns and structure in this data and its transformation into actionable insights is driven by applications that use AI/ML/DL workloads.
|Vexata VX-100F Specifications|
|Technology||Fibre Channel attached NVMe|
|Connectivity||Up to 16 x 32Gb/s Fibre Channel|
|IOPS*||Up to 6 Million IOPS|
|Throughput*||70GB/s (50GB/s R; 20GB/s W)|
|Latency*||100% Write: 50us
100% Read: 105us
|Usable Capacity (3.84TB)||175TB (RAID 6) - 187.5TB (RAID 5)|
|Usable Capacity (7.68TB)||406TB (RAID 6) - 425TB (RAID 5)|
|Rack Space (RU)||6|
|Max Power Consumption||2900W (2+2)|
|OS Support||Linux, Solaris, AIX, HP-UX, Windows, VMware vSphere|
|Management||Web UI, CLI and RESTful API|
|Resilience||Active-Active High Availability Design, RAID 5 or RAID 6, Hot Swappable ESM blades, Non-Disruptive Upgrades for ESM and IOC modules, Non-Disruptive Capacity and Performance Upgrades|
|Data Services||Thin Provisioning, Space Efficient Snaps and Clones, Pattern Removal, Always-on Encryption with no performance impact|
* Measured under load using multiple test cases (such as VDBench, FIO, Oracle SLOB2)
Front Cover View
Download the Vexata VX-100F Scalable NVMe Flash Array Datasheet (PDF)
Contact us for pricing!