Edge will provide a more energy efficient power source for emerging tech
By Michał Kijewski, solutions manager at Mobica
Almost a hundred years ago the world witnessed a huge leap in industrial output. The emergence of the electric grid resulted in plummeting energy prices – kick-starting the mass production revolution in the automotive sector.
The impact was huge. For example, at the start of the 1920s the average car in the US cost $940, but by the end of the decade it was just $290.
When it comes to such industrial advancements, analysts are still optimistic that, despite the impact of the pandemic, the 2020s will roar even louder than their twentieth century counterpart.
The reason behind this positive sentiment is that we once again stand on the cusp of major technological innovation. The emergence and convergence of 5G, IoT and AI will lead to more devices and people being connected to rapid computing power. For instance, 5G promises speeds up to 100 times faster than 4G, with extremely low latency, and there is little doubt this will unleash enormous potential.
There have been some outlandish forecasts around what the result of this development will be – predictions range from flying taxis to holographic phone calls. It makes for interesting reading, but the shorter-term realities may be slightly more restrained.
This is not to say they won’t be revolutionary though. While in the 1920s, technological advancements made vehicles affordable to the everyday man, in the 2020s these developments will create the fully connected car. Enhanced by dozens of sensors, cameras, apps and entertainment systems, the automobile is now beginning to resemble a smartphone on wheels.
The power struggle
These incredible capabilities come at a cost, however, in the form of energy demands. For example, a connected car using multiple advanced driver assistance systems (ADAS) requires powerful AI – to process and analyse all the data coming from various sensors. When a crash is imminent, the ability to do this in microseconds can mean the difference between life and death.
When we are asking cars to process that much data that fast, it will take a toll on the energy supply. If we are going to avoid draining the vehicle completely, we are going to need to offload as much data processing as possible into the cloud.
Cars are not the only machines that are looking to offload their power demands, however. IoT devices, enabled by the rollout of 5G, are increasing cloud data centre workloads exponentially. Cisco predicts that the number of devices connected to IP networks will be more than three times the global population by 2023.
The energy demands required to transport data back and forth between devices and a core data centre has become significant. To combat this energy consumption and to reduce the latency involved in the transportation of data, we’re now seeing the emergence of edge computing – a more distributed approach to cloud computing.
Edge computing is creating an infrastructure that will allow data processing to take place away from a core cloud data centre and closer to the device. Just as the industry was transformed by the creation of the electric grid in the early 20st century, we are likely to see a cloud computing grid be equally transformative over the next decade.
Product designers are now having to give special attention to what workload should be processed on devices, and what should go to core cloud data centres, and what should go to the edge.
There will always be a trade-off between the need for compute power and energy supply. But, as edge computing becomes more prevalent and emerging technology can interact with a localised cloud infrastructure more readily via 5G, we will be able to manage those energy demands in a more efficient way.
This article first appeared in the April/May 2021 edition of Industrial News: Electronics & Engineering.
