Far from being a clean sector, IT is an immense consumer of natural resources and electricity. The majority of the world’s semiconductors are manufactured in Taiwanese fabrication plants (fabs), which Greenpeace estimates, consume 12% of the country’s total electricity. In 2019, it warned that TSMC, the world’s largest chipmaker, whose customers include Apple, Google and HP, only powered 5.4% of its operations with renewables.
Fabs also use a huge amount of “ultra pure water” to clean the wafers used in manufacturing. In 2015, an article in Harvard Business Review reported that a single manufacturing plant can use 2-9 million gallons of water a day. The wastewater from semiconductor fabrication is also highly toxic and there is a significant carbon footprint in the journey of a chip as it moves through a manufacturing supply chain to end up as an item of IT equipment that is ultimately shipped to a customer.
When the equipment has reached the end of its life, there is the environmental impact. Clearly, once the equipment is up and running, it has a useful life running software, which may help in lowering its carbon footprint. But software uses processor resources; the more processing required, the more electricity is consumed.
Last year Wired quoted Evan Sparks, CEO of Determined AI, who estimated that OpenAI’s machine for solving the Rubik’s Cube puzzle, consumed 2.8 gigawatts of electricity. This is roughly equivalent to the output of three nuclear reactors. While AI will increasingly have a role in society, some experts are calling for a regulatory framework that legislates against excessive power consumption. With more computational workloads consuming cloud-based IT resources, there is also growing pressure for the major cloud providers and web giants to become more accountable on the sustainability and environmental impact of their vast datacentre operations.
Digitising for the circular economy
IT clearly has an environmental impact. But it can also improve sustainability. A key benefit of digitisation is that it encourages seamless data flows. Much of the focus has been to patch broken internal processes. But there is a huge opportunity in data flow through a supply chain.
A supply chain tends to flow linearly. Tracking of waste materials is not handled in the same way as the procurement of raw materials. But somebody’s waste product is another organisation’s raw materials. By linking this data flow, it is possible to build a circular economy, which not only improves sustainability, but also facilitates greater supply chain resiliency. As the world turns its attention to COP26, there’s an opportunity for everyone to improve sustainability and carbon reduction efforts.