Steve Wozniak, the man who singlehandedly engineered the Apple I and Apple II computers that launched the personal computing industry, thinks that virtual storage is the next big thing.
The Apple co-founder is named chief scientist at Primary Data just shortly after the Silicon Valley startup came out of stealth.
At the Demo Fall 2014 conference in San Jose, California, Wozniak joined his former Fusion-io colleagues on stage to make the announcement. Fusion-io was founded by Wozniak, where he was also chief scientist, with then chief technology officer now Primary Data CEO Lance Smith. Fusion-io co-founder David Flynn is also part of Primary Data as the company's chief technology officer.
Wozniak believes more and more enterprises will need faster and better ways to store and deliver data as consumers increasingly adopt new technologies such as wearable devices, personal digital assistants and Internet of Things devices. His role in Primary Data will be to help develop the technology that makes virtualization in data centers a mainstay while advocating for the company's products as it attracts new clients.
"Primary Data uses data virtualization to dynamically place data anywhere as applications demand," says Wozniak. "I'm excited to share our data virtualization platform with the world."
Virtualization in data centers is not a new concept, with companies such as VMWare and EMC having developed tools that allow the virtualization of server systems and, more recently, networks that will allow multiple applications to run simultaneously.
Storage, however, remains constrained because data is often limited by the type of storage hardware that contains it. Flynn says the different types of hardware used in data centers create separate silos that slow down data retrieval and delivery.
Primary Data's goal, however, is to deploy its custom-built software based on the widely used Network File System that will convert metadata to software code and allow the hypervisor to quickly track and route the data to where it is requested.
Flynn says Primary Data's software can interpret the protocols of each storage hardware, whether it be flash, optical or other, and this is why the type of storage hardware doesn't matter. It treats each one of them as if they were all one and the same, allowing users to better allocate storage resources and retrieve data more efficiently by running the hypervisor.
"This allows us to scale across different types of storage and get a level of performance that you don't see today," says Flynn.
For now, Primary Data is offering its product as an appliance that needs to be hooked to clients' existing storage systems, but the startup is working on a software version that can be delivered through the cloud.
A few customers, mostly media companies who have large amounts of data to store, have already jumped on board as testers before Primary Data officially launches its virtualization tool by mid-2015.