SUSTAINABILITY
The engineering case for sustainable electronics design is easily made
By Taylor Aston-Nielsen, SiTime
Is sustainability a constraint on engineering innovation, or a result of it?
The pressing need to reduce energy consumption is being ratcheted up by government targets such as Clean Power 2030 and Net Zero 2050. But the goal of making sustainable products is achieving more than merely falling in line with regulations. It is paving the way for a generation of electronics with unprecedented power efficiency, accuracy and resilience.
Engineers are increasingly designing for resource optimisation and longevity from the outset, and the result is a wave of products that are smaller, consume less and can operate in harsh environments. In this sense, the acceleration of global electricity demand – a 50 per cent jump in average growth pace expected from 2026-2030 – driven by the exponential expansion of data centres and AI, has ushered in a new era of innovation.
Search for efficiency amid the AI boom
Nowhere is the evolution of sustainability from a compliance obligation into a core engineering design principle is this more evident than in the growth of AI infrastructure. The top six technology giants will plough a projected $700 billion into data centres in 2026 alone, but GPU utilisation in AI clusters is as low as 35-45 per cent of potential performance. That’s why system efficiency in the form of performance per watt (PPW) has become a critical metric. PPW has emerged as a metric that cleanly connects engineering ambition to environmental outcome. In the data centre business, where energy costs and grid constraints are intensifying, PPW has become a standard unit of competitive comparison. In industrial and consumer applications, it is related to everything from battery life, thermal management
28
“
Better engineering
and better environmental performance are increasingly the same objective, pursued through the same design decisions.
requirements and total cost of ownership. The AI scaling challenge makes the stakes concrete. Training large-scale models requires thousands of synchronised processors at high utilisation. As AI scales, so does the energy it consumes. Improving PPW requires component-level decisions with system-wide consequences to be taken correctly.
Multi-dimensional
component-level decisions One of the more consequential shifts in sustainable design thinking is the recognition that system-level efficiency is
SEPTEMBER 2026 | ELECTRONICS FOR ENGINEERS ”
built from component-level decisions. This requires a change in how components are specified. We are moving away from standalone specifications to assessing a component’s contribution to system-wide power consumption, component count and lifetime performance.
Beyond operational power, component- level innovation is shrinking BOMs. Higher integration is combining functions which had previously required multiple discrete parts into a single device. This reduces total component count while saving manufacturing energy and materials. In the harsh, high-density conditions of AI datacentres - as well as the variable environments of IoT networks and other connected products - components with a greater resilience to shock, vibration, and temperature variation are increasingly indispensable. Operationally, resilient components perform better under environmental stress. They also eliminate the need for redundant backup devices and reduce field repair visits, lowering the fuel consumption and emissions associated with
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50
