Actu
Violin Memory helps deliver Cisco’s world record performance on Cloud Computing Benchmark
September 2012 by Marc Jacob
Violin Memory, Inc., provider of a new class of high-performance flash-based storage systems, today announced that it served as the storage platform that helped to enable the new world record 8-node VMmark 2.1 score of 42.79@36 tiles. The benchmark was delivered by the Cisco Unified Computing System (UCS) B200 M3 Blade Servers. A record of this magnitude validates the sustained performance achieved by Violin Memory Arrays and their ability to provide the infrastructure necessary to support virtualised and cloud environments.
VMmark 2.1: The Industry’s Virtualisation Platform Benchmark
VMmark 2.1 is the industry’s first multi-server datacentre virtualisation benchmark, which assesses the performance of a group of virtualised real world applications. It includes a variety of common platform-level workloads such as live migration of virtual machines, cloning and deploying of virtual machines, and automatic virtual machine load balancing across the datacentre. To achieve the best results, end-to-end performance from server to storage is needed. The benchmark demonstrates virtualisation and infrastructure performance as well as the agility required for cloud computing.
Violin Storage at the Speed of Memory
The VMmark 2.1 results demonstrate that the Violin 6000 Series Flash Memory Arrays deliver the industry leading IOPS and latency performance required to virtualise and consolidate demanding business critical applications and desktops. Details of this record-setting result can be found here. This result is followed with the updated dual node 2-socket record set by Cisco, earning a score of 11.32@10 tiles using VMware vSphere 5.1. Here is the VMmark disclosure document.
The Violin 6000 Series Flash Memory Arrays bring storage performance in balance with high-speed compute and networking, offering a significant opportunity for infrastructure consolidation. A single system fits in 3U of rack space and can deliver up to one million IOPS with 4 GB/sec of bandwidth – enough performance to replace multiple racks of traditional disk arrays for savings of both CAPEX and OPEX. The arrays attach to the network for shared primary storage. Multiple arrays can be clustered together to achieve petabytes (PB) of capacity and high aggregate bandwidth.
