Posted: 29 May 2019 10 min. read

Machine learning – not a chip off the old block

Earlier this year we announced our TMT (technology, media and telecoms) predictions for the UK tech and digital sectors in 2018. These annual predictions are published in full in the 17th edition of Technology, Media & Telecommunications (TMT) Predictions 2018.

The predictions are based on hundreds of conversations with industry executives and tens of thousands of online interviews with members of the public around the world.

This blog focuses on the predictions around machine learning applications and next gen machine learning chips.

 

Chipping in!

The impact of machine learning applications – from machine vision to voice recognition, will be massively enhanced via the deployment of next generation machine learning centric chips.

Machine learning is an artificial intelligence, or cognitive, technology that enables systems to learn and improve from experience—by exposure to data—without being programmed explicitly.

Our prediction indicates that large and medium-sized businesses will likely double their use of machine learning technology by the end of 2018.

The report highlights five key areas that we believe will unlock more intensive use of machine learning within organisations by making it easier, cheaper and faster.

Next generation chips

The most important key area is the growth in new semiconductor chips that will increase the use of machine learning. These next generation chips will enable applications to use less power, and at the same time become more responsive, flexible and capable.

A surge in deployment of FPGA (field programmable gate array) chips and ASIC (application-specific integrated circuit) chips will be a core driver in the use of machine learning. These new kinds of chips will represent between a quarter and a third of all machine learning chips deployed this year.

Machine learning has previously been reliant on GPU chips (graphics processing units), which may be slower or more power hungry than FPGAs and ASICs.

It isn’t just the chips that are getting better. Automation, data reduction and training will make machine learning technology easier, cheaper and faster for businesses to adopt and implement across a variety of sectors. In 2018, machine learning will become increasingly mainstream – but there will still be plenty of progress to make in subsequent years.

Want to find out more?

Duncan Stewart, director of TMT research at Deloitte, talks about the prediction for machine learning in more detail in this video:

Machine Learning | Deloitte TMT Predictions