The future of storage

With the storage industry being touted as one of the few sectors able to survive the current economic crisis, and with budgets relatively unscathed, it is interesting to explore what the future might hold for this sector.

We will certainly witness the usual advent of disk sizes and speeds doubling rapidly - but due to current hard disk drive limitations, we will soon need new technologies, and there are some radical ideas to address this explosion of digital content.

Let's take a look at some of the interesting ideas out there.

One proposal being bandied about is the spinning mechanical platter being replaced by a storage device that oscillates.

This device would offer higher capacities and performance and would not suffer from the fatigues that cause existing spinning disk technologies to fail. This technology would also be less susceptible to damage that can affect the current mechanical devices, which are unable to handle G-forces well.

Due to the nature of its design, the device won't rely on a mechanical motor, so there would be less friction and heat generated. This in turn means less power is needed to drive and cool the device, keeping the green movements happy as well as reducing costs.

While the hype surrounding the solid state disk (SSD) suggests we need look no further for alternatives, costs for this technology are still prohibitive. Also, the device has a somewhat limited lifespan for reads and often doesn't provide the capacities required for today's needs.

Further ahead and still theoretical at this stage - are ideas of holographic storage. While this may seem like science fiction, we need major advancements in technologies to reach the performance and storage densities we need in the digital media.

So what is holographic storage?

It would, in theory, no longer be applying a two-dimensional view to storing data, but move into storing data in a three-dimensional format using holograms (holographic cylinders using nanotechnology).

To achieve a functioning storage device, a 3D grid will be read by multiple lasers simultaneously, and allow storage of multiple terabytes in a form factor offered by a CD; this is possible due the density of information that could be stored in a 3D space.

Other benefits would come from the nature of the design of a 3D grid, including speed; the device would be accessed by multiple lasers without having moving parts. This would offer instant data access for multiple streams concurrently, allowing improvements in speed of data delivery and the ability to provide multiple input/output operations simultaneously. Early reports suggest we could be looking at 100TB of capacity.

Some other futuristic devices flaunted as the new frontiers of data storage include molecular memory. A research team at a US university has determined that a strip of graphite only 10 atoms thick can store data, making vast amounts of storage available for computers, cellphones and cameras in miniscule spaces.

The devices are rugged and would consume little power and generate almost no heat. This could promote huge advancements in the flexibility of mobile devices and allow for a high specification laptop that does not overheat, with a battery that lasts for several days when in use.

Further ahead and still theoretical at this stage - are ideas of holographic storage.

Adam Day is product manager at SYSDBA.

These ideas start to push the boundaries of traditional computing, as data access from disk is at present the lowest common denominator in the chain and a large amount of design goes into caching data, whether in the disk or memory, to allow CPU cycles to continue uninterrupted.

With the advent of improved technologies in one area, we could see advancements in many other areas as a result.

There have been past attempts to advance the industry, which have failed dismally - have a look at the following blog to see some innovations that were tabled and disappeared nearly as quickly as they arrived: http://blog.fosketts.net/2008/12/06/top-ten-coolest-enterprise-storage-flops/.

Let's hope these new technologies make it into production in the near future, as the disk drive has been the slowest link in the computing chain for some time. While it has been a faithful friend, the requirements to store and access data grow in double digits every year. Yet we are still reliant on a device that has not fundamentally changed since the early days of computing.

IBM introduced hard disks in 1956 and, in truth, the architecture has changed little over the years - it is still a mechanical device. So let's push the boundaries to come up with something fantastic, fast and inexpensive.

That way we can afford to digitise our DVD collections and continue to snap away with our digital cameras without having to invest in an enterprise array, and hopefully allow the corporate out there to finally do more with less, as this is what we have been asking for all along.

* Adam Day is product manager at SYSDBA.