A look at how storage technologies have changed since the days of punch cards and magnetic tape.
Companies are capturing more data than ever, and the demand for storage is growing as the industry learns how to do more with data. According to IDC, 34.7 billion gigabytes of storage were shipped in the second quarter of 2016 alone. Compare that to the very first hard-disk drive that was produced in 1956 – which came in at five megabytes of storage – and consider the fact that it would take seven trillion of those disk drives to match what the industry produces now.
The world’s appetite for storage is continuously growing thanks to the number of devices that are generating data, which is contributing to the warehouses of information kept and classified as big data. The growth in big data is fueling innovations in developing cognitive learning applications, such as artificial intelligence, machine learning, and big-data analytics. This in turn is creating a world that is more predictive, productive, and personal because it enables people to make more informed decisions.
Meanwhile, innovations in storage such as solid-state drives have helped to make tasks that were previously impossible - due to performance bottlenecks - entirely possible. For instance, as social media has grown rapidly to accommodate user bases that number in the hundreds of millions, the demand for updated, real-time information has created latency problems that have been solved and reduced by three-fold thanks to the performance capabilities of SSDs.
Object storage on the other hand, enables public and private storage clouds to manage data at enormous scale. Amazon’s S3 protocol has become the accepted industry standard because it enables access to data anywhere, at any time, across any device. Object storage is able to achieve this voluminous comprehension of data by scaling out instead of scaling up, which can accommodate any required level of storage capacity.
While SSDs and object storage are the latest trends that are transforming storage now, what may be surprising is that this evolution of storage goes back much farther than just the last few years. Let's look back at ways data has been stored, with innovations dating back to the 19th century, as well as storage technologies that are propelling us into the future.
Early forms of data capture
Before the age of personal and supercomputers, filing cabinets were used to store all office data. In the late 19th century, data was punched into cards that could be interpreted by tabulating machines known as unit record machines. Originally used to record results from the 1890 census, these machines could quickly store and process data, much faster than a human could. And as the process became automated, the ability to store more data and process that data faster led to leaps and gains in terms of understanding and manipulating data.
(Image: US Census Bureau employee transferring handwritten information to a punch card using a card puncher in 1930. Source: US Census Bureau)
As the amount of data that could be “stored” on punched cards became physically limited by the space on the card, magnetic tape storage became a popular replacement in the 1960’s and hit peak adoption in the 1980’s due to its low cost and effectiveness at storing larger quantities of data. Today, tape storage is still in use, but is seen as being useful as “cold” storage for data that does not need to be accessed immediately.
The hard disk drive is born
The IBM 350 RAMAC was the first hard disk drive, and it stored up to five megabytes of data on 50 24-inch disks. The entire unit weighed approximately one ton. At the time of its release in 1956, the drive could be leased for $3,200 per month. The first customer to lease one of the drives was United Airlines, who used it to store data from its reservation system.
Hard-disk drive breakthroughs
Over the past 60 years, since the IBM RAMAC was first produced, hard-disk drive manufacturers have used new technological breakthroughs to reach higher capacities, better performance, and sensible form factors for use in servers and data center system racks. Most recently, helium-filled drives have pushed capacity limits to create denser, energy-efficient drives that drive down total cost of ownership, allowing storage arrays to hundreds of terabytes. Advanced systems approach multiple petabytes and offer capacity that can continue to scale well beyond current data needs.
SSDs drive performance
Flash storage has enabled applications to run at higher peak performance while efficiently utilizing personal and server compute power to run more applications at once. Recent breakthroughs in using non-volatile memory (aka NVMe) for solid-state drives have increased the performance gains to the level of DRAM, and show how much promise flash has for performance-based computing that enable fast data.
Object storage powers the cloud
Object storage is a fairly new innovation that challenges traditional file-based storage as it is better suited for massive capacity environments. Object storage is the prevailing technology behind some of the largest cloud providers because it offers scalability to any capacity, provides a single pool of storage, can be geo-distributed to ensure that data is always available, and leverages low-cost and energy-efficient hardware to create a lower total cost of ownership.