Amplidata: Redefining Massively Scalable Storage for Big Data

The buzz about Big Data is getting louder as the mountain of data continues to double every two years. Businesses are challenged with how to efficiently store, protect and mine the data for a competitive advantage. Mike Wall, CEO of Amplidata believes they have the answer for Big Data storage. Founded in 2008, the company recognized that traditional storage based on RAID was no longer sufficient for Big Data and a new approach was needed. Using proven state-of-the-art rateless erasure code technology from the telecommunications industry and a great deal of secret sauce, Amplidata developed AmpliStor®; an object-based software defined storage solution that delivers exceptionally high levels of data durability, scalability and operating efficiency at a much lower price than traditional storage solutions.

The challenge with RAID for longterm Big Data storage is that large multi-terabyte SATA drives are used to be cost effective. When a large drive fails, the rebuild time can take weeks, significantly increasing the risk of data loss if other failures occur. System performance also takes a big hit during the rebuild making it difficult for users and applications to do their jobs.

AmpliStor has addressed the RAID and other large-scale storage challenges with its fully abstracted software stack optimized for Intel-based commercialoff-the shelf hardware. At its core are patented BitSpread® and BitDynamics® technologies. Bitspread can achieve fifteen 9s data durability tolerating up to 19 simultaneous drive failures. BitDynamics continuously monitors data integrity and automatically self heals with virtually no impact to system performance. Another key technology is GeoSpread that spreads data across multiple data centers for disaster prevention using 65 percent less capacity than RAID6 plus replication. With GeoSpread, full data protection and access are maintained even if an entire data center becomes unavailable.

Scalability is flexible and virtually limitless to meet varying customer needs. Performance and capacity scale independently and dynamically by simply adding controllers or storage nodes. A close relationship with Intel means AmpliStor is fully optimized to get the most out of the latest processors. Throughput performance is exceptional and scales linearly with each controller. The base configuration of three controllers delivers three gigabytes per second throughput and can be scaled to tens of gigabytes per second. Capacity is easily scaled from hundreds of terabytes to exabytes by adding storage nodes.

With support for common protocols including NFS, CIFS, HTTP/REST and S3, AmpliStor integrates well with existing environments. A global namespace makes the same data accessible by teams and applications wherever they are located. Manageability is made easy with high levels of automation. AmpliStor can be scaled or reconfigured on-the-fly without having to take the system down or migrate data. The system will automatically and dynamically rebalance to the new durability policy or configuration, ensuring data is always protected and accessible.

“AmpliStor is ideally suited for companies looking for cloud-scale economics with the security, flexibility and performance of their data center,” said Mike Wall, Chairman and CEO of Amplidata. “We are working with some of the largest companies in the world who recognize the need for a different approach to Big Data Storage and how our solution helps solve their challenges.”

AmpliStor is ideally suited for companies looking for cloud-scale economics with the security, flexibility and performance of their data center