What is racetrack memory?
This experimental technology could change the face of how we store data
As storage bursts at the seams with people and organisations storing ever-increasing amounts of data, researchers may have technology that could store hundreds of times more data than today’s best efforts.
Called racetrack memory, the tech is also known as domain-wall memory (DWM). It is a tentative technology under development that's projected to offer a replacement for current memory types, including hard disk drives (HDDs) and flash.
It is a type of non-volatile, solid-state memory. However, its inventor, IBM, hopes will one day be able to hold 100 times the amount of data that can be stored on any technology that is commercially available at present.
"The technology could have the potential to enable handheld devices to hold a few thousand movies, run for weeks at a time on a single battery and be practically unbreakable," IBM claims.
Drives with the technology could also be cheaper than flash memory, IBM hopes it will cost the same as traditional hard disk drives. Therefore it's said to bring together "the best of both worlds" of the primary memory types in use today.
How does racetrack memory work?
Racetrack memory uses spintronics - the inherent strength and orientation of the magnetic field produced by an electron as it spins - in addition to its electronic charge, in solid-state devices. Solid-state in this sense refers not only to the kind of storage we're familiar with in the market today, but also older technologies like vacuum tubes.
As an area of study, spintronics has been around since the mid-1980s, but it's only been put into any practical use in the past 15 or so years. It wasn't until 2007 that IBM researchers first published a paper on racetrack memory in American Physical Society's Physical Review Letters. This was followed by a further paper in the journal Science.
As described in the abstract of the 2008 paper, racetrack memory "comprises an array of magnetic nanowires arranged horizontally or vertically on a silicon chip". These nanowires, measuring around 50nm in diameter, form the so-called racetrack.
To use a metaphor coined by Stuart Parkin, one of the pioneers of racetrack memory, the nanowires are like skyscrapers with each floor of each skyscraper containing a single bit of data. A transistor at the bottom of the wire shoots spin polarised electric currents up and down the wires, which moves the data up and own.
As with other electronic memory, it uses reading and writing heads located near the storage medium in this case, the nanowires which then cause elements of the medium to lie one way or the other, encoding the data onto it in binary form.
Where it differs is that it doesn't write just to a single side of the storage medium. Instead, it writes on the whole of the "domain wall" of the nanowires, making it one of the first examples of 3D data storage.
This massively increases the quantity of data that can be held on a single wire.
"In this way, each transistor can store not just one bit of data, as in all other solid-state memory, but rather 100 bits," Parkin said in a 2011 profile. "This means that one can have a solid-state memory with the same low cost of a disk drive but with a performance 10 million times better!"
Join us for the inaugural IT Pro Live conference June 22-26
Stream it live from the safety of your own home or officeRegister now
What are the advantages of racetrack memory?
The primary advantage of racetrack memory is how cost-effective it is. While the price of solid-state memory has decreased markedly in recent years, traditional spinning disk hard drives are still far cheaper.
Because racetrack memory can pack so much data onto single nanowires, per transistor it's a cheaper form of storage than current solid-state memory technology, hence the comparison with HDDs.
It's also much, much faster to read and write data than on any extant technology. The researchers behind the technology predict it will be able to read or write a bit of information in 10 nanoseconds or less. By contrast, a hard disk would take around three million nanoseconds to do the same.
What are the disadvantages of racetrack memory?
Despite it being around for some time now, racetrack memory is still considered an experimental technology. This is because working at the nanoscale while also trying to tackle some of the whims of quantum physics comes with certain obstacles.
One of the encountered challenges is the issue with domain walls - they can move or stop at random, with no definitive way of preventing this from happening.
Another common challenge is that even the smallest imperfection in the wire (“small” is defined as "detectable only with an x-ray microscope") has the power to significantly reduce read/write times to the point that the performance is closer to that of an HDD.
When can I buy a racetrack memory storage device?
Answering this question is not as straightforward as you would expect. According to Parkin’s most recent estimates, you might be able to purchase a racetrack memory storage device in around five years, but only if there’s sufficient investment prior to that time.
However, there have not been any further updates since 2007 and a few of the above-mentioned issues are yet to be overcome. Unfortunately, unless someone produces a solution preventing domain walls from randomly moving, there's not much chance of memory storage devices hitting the market anytime soon.
On the other hand, taking under account sufficient support, racetrack memory has the potential to revolutionise the way we utilise and experience data storage as well as retrieval. However, we might just have to be patient and hope for the best.
Security analytics for your multi-cloud deployments
IBM Security QRadar SIEM solution briefDownload now
Five reasons to move to the cloud
Join the enterprises moving their workloads to the cloudDownload now
Architecting hybrid IT and edge for digital advantage
Why business leaders should consider a hybrid IT strategyDownload now
Six reasons to accelerate remote asset monitoring with AI
How to optimise resources, increase productivity, and grow profit margins with AIDownload now