Analysis
Marvell's New DragonFly Virtual Storage Accelerator Introduces Breakthrough Host Cache Architecture for Enterprise Storage
Marvell (Nasdaq: MRVL), a worldwide leader in integrated silicon solutions, today launched the Marvell® DragonFly™ Virtual Storage Accelerator (VSA) powered by HyperScale™ embedded processor technology.
In wThe DragonFly VSA represents an entirely new architecture in the industry, achieving the benefits of virtualization scaling and enterprise NAS/SAN/DAS storage cost reduction via turn-key hardware.
The Marvell DragonFly VSA is a revolutionary product for enterprise-class cloud computing, offering powerful host server I/O cache acceleration to solve worsening storage I/O bottlenecks, said Pantas Sutardja, Chief Technology Officer at Marvell. The HyperScale caching technology embedded in the DragonFly VSA is the product of leveraging our leadership in HDD, SSD storage technologies, and extending our advanced-integration capabilities. The DragonFly VSA presents itself as a powerful tool for data center architects in optimizing the costs of the server/storage infrastructure. We look forward to collaborating with visionary Cloud storage architects to achieve a leapfrog improvement in virtualization scalability.
Addressing the issue of I/O bottlenecks in the enterprise is a real problem facing many data center managers, said Jeff Janukowicz, research manager for Solid State Storage Technologies at IDC. A turn-key system solution, such as Marvell's DragonFly Virtual Storage Accelerator, that combines silicon with additional hardware in an industry standard PCIe card that addresses these performance concerns is an innovation that will be welcomed by many enterprise storage customers.
Marvell is changing the game for organizations that have been faced with implementing multiple application servers to effectively run email, databases, warehousing and much more in a virtual or cloud environment, said Arun Taneja, founder and president, Taneja Group. DragonFly VSA is moving performance up to the application server itself instead of relying on storage for the I/O cache. It simply plugs into the server and fits in seamlessly with existing architectures to provide a tremendous boost in systems efficiency and performance. It also releases the pressure on the backend storage to carry the entire load for IO yet uses the capacity-aspects of storage very effectively.