Historically, power and cost have been seen by electronics engineers as imposing barriers to the implementation of AI into end point devices – appliances that communicate and cooperate to form part of the Artificial Intelligence of Things (AIoT).
Typically, the processors capable of enabling this would have come at a high cost to the developer, having been designed for very specific purposes on the most advanced process nodes. This high-performance tech came at a logistical and environmental cost too, consuming significant amounts of energy and requiring a lot of power to run.
The silver lining
Yet the clouds have cleared, it seems. According to the results of our third annual report exploring engineer’s attitudes towards AIoT capabilities, the tide is shifting, and the possibility of AI-optimised edge technology is becoming more of a reality for most people in the field.
Our Edge of Now survey found that just 24% of engineers now see cost as a barrier to increasing the on-device processing in their designs. That’s half of the 48% from last year’s report, and down from 64% in 2020.
Processing capacity is less of a problem too: barely a tenth of engineers (13%) see power requirements as a blocker to introducing AI to their products. With increased processing laying the groundwork for onboard AI, this is huge for the development of the AIoT market.
Out of sight, out of mind
While it’s encouraging to see engineers so positive in tackling these obstacles, they haven’t completely disappeared. We should view a surge in positivity and ambition in the industry as the perfect time to push for the elimination of the lingering challenges to the adoption of the AIoT: to harness more onboard processing capabilities, allowing for more sophisticated AI functionality to be added to end point devices.
But to build a reality where product developers and engineers can maximise the potential of their creations in this way, we need to continue to enable innovation. Demand is rising for flexible silicon solutions that will make way for true customisation and control. In turn, driving down costs, combating high power consumption and paving the way for real AIoT progress.
The smart solution for smarter tech
XMOS’ contribution to this vast market need is xcore®.ai: our disruptive crossover processor designed to deliver flexible, high-performance AI, DSP, control, and IO in a single chip.
xcore.ai enables data to be processed locally – and actions to be taken on device – within nanoseconds, and without communication with the cloud. Encompassing the capabilities of both an applications processor and a microcontroller, it delivers seamless results without added complexity, energy consumption or cost.
As the industry gears up for a sea change of growing appetite and readiness for the commercial adoption of the AIoT, we are excited to be playing our part, enabling the development of new applications across a fragmented market. Implemented on a mass scale and with empowered, educated designers, developers, and manufacturers the possibilities are endless!
You can catch our insights in this year’s Edge of Now report, here.