Bladewatch.com

Business Wire

AUSTIN, Texas–(BUSINESS WIRE)–NextIO (www.nextio.com), the premier provider of next-generation server I/O, will demonstrate its second-generation I/O virtualization (IOV) solutions for the first time at the 2009 SIFMA (Securities Industries and Financial Markets Association) Technology Exhibit and Conference in New York City, booth #2205, June 23-25. These solutions are optimal for Enterprise Data Center and High Performance Compute customers needing a mixture of flexibility, low cost, and ease of management.

NextIO is the industry leader in providing intelligent, rack-optimized I/O virtualization products based on high-performance PCI Express (PCIe) switching technology. NextIO’s Express Connect™ products enable I/O controllers to be virtualized and shared across physical servers in a variety of form factors, including blade and rack servers. The result is an unprecedented combination of low cost, low power, and flexible I/O with simple, easy to use management.

NextIO’s Top-of-Rack product for Enterprise Data Centers deploying Ethernet and Fibre Channel will be used to showcase the following use cases:

• Rapid provisioning of servers in minutes, compared to the days or weeks required today. Servers utilizing NextIO can easily be pre-provisioned for fast install or re-provisioned based on demand.
• Fast N+N failover recovery in an N+1 architecture, with no impact to the network. NextIO delivers all the performance of an N+N server framework without the expense.

NextIO will also highlight its recently announced partnerships with Marvell and NVIDIA for HPC customers. NextIO and Marvell will demonstrate a storage solution that can achieve over 200,000 IOPs and 400GB per PCIe slot. The solution scales to well over one million IOPs and 4 terabytes of SSD storage in 3U, using less than 500 watts of active power. NextIO will also demonstrate how its partnership with NVIDIA accelerates CUDA development and deployment using NextIO Adaptive Connect^® PCIe chassis and management software. The demonstration will showcase from 1 to 5 NVIDIA GPU’s connected to a single server and compute the performance of each configuration using a Monte Carlo financial trading simulation application.

Anything we can do to improve the performance of i/o whether it’s for that market trading system, or for low latency hpc solutions has to be a good thing for end user choice and platform possibilities, I’m off to read up more.

No related posts.

Related posts brought to you by Yet Another Related Posts Plugin.