The energy impact of electronics has multiple layers and must be considered holistically.
Sustainability has become a “scorching topic” (pun intended) in discussions I have with customers and ecosystem partners. It is now vital to many corporations as part of what many they report on “Environmental, Social, and Governance” (ESG). It can also be very confusing. In this post, I am trying to clarify some confusion after attending and organizing several related events recently.
Attending the Xtreme Technology Challenge (XTC) finals at “TechCrunch Climate” as a sponsor in June gave quite a fascinating insight into related technologies. It was fun to see Bill Gates live and provide guidance on what climate technologies he’s looking to fund. He pointed to industrial, agriculture, and land use activities, some of which are his Breakthrough Energy Ventures fund addresses. He asked for more ideas. But he also predicted somewhat of a “Climate Tech Winter.”
One big takeaway from the event, including an interview with Impossible Foods founder Patrick Brown, was the existing confusion regarding the significant contributors to climate change. You can find some interesting background reading in Dr. Jonathan Foley’s article “The Three Most Important Graphs in Climate Change.” CO2 from fossil fuels, land use, and chemicals makes up 76% of greenhouse gas emissions, followed by 16% methane and 6% nitrous oxide. The small remaining balance goes to so-called f-gases, mainly hydrofluorocarbons, chlorofluorocarbons, and other fluorinated gases. As Foley states, “Globally, the two biggest sectors that contribute to climate change are electricity generation (~25%) and food & land use (~24%). In other words, burning coal, oil, and natural gas to generate electricity is the single largest source of global emissions, but the food & land use sector is nearly tied with it.” Industrial use accounts for 21%, transportation for 14%, and building for 6%.
Land use is often a surprise. The three main drivers in order of importance here are deforestation for food production, methane production by cattle and rice fields, and nitrous oxide emissions from overusing fertilizers in agriculture.
So what is the impact of electronics, and by extension, the impact of EDA?
On a personal note, my friends and family circle sometimes say, “Well, the data centers that you guys enable have such a significant impact.” This, of course, does not prevent immediate commentary on why our industry does give them sufficient 5G coverage or bandwidth while being on the road in Berlin in London. Oh well. Spoiler alert – blaming data centers by itself is not appropriate. As outlined previously here and here, from 2010 to 2018, improved efficiencies in electronics and their design have limited the data center energy growth to only about 6% while delivering 6x more workloads, 10x more internet traffic, and 25x more storage. That said, consumer demands (including those of my friends and family) are and will always be insatiable.
To address these concerns, I have been following some of the industry predictions for a while. The research “New perspectives on internet electricity use in 2030” by Anders S.G. Andrae from 2020 provides the latest predictions. Taking Andrae’s source data on the consumption of “Information Technology” and adding them up shows the graph below.
The biggest contributor clearly is the device side, followed by industrial production and data centers, which are about comparable to the sum of fixed and mobile networks. And bear in mind that this is only the direct impact these IT components consume, but it is important to consider the scope of impact of electronics.
When considering the “scope” of impact on emissions, there are existing industry definitions out there. Per PwC, Scope 1 emissions are direct emissions from owned and controlled sources, such as emissions associated with fuel combustion in boilers, furnaces, and vehicles. Scope 2 includes indirect emissions from purchased electricity, steam, heating, and cooling. The bucket of Scope 3 emissions includes indirect emissions that occur in a company’s value chain by customers and/or suppliers, such as emissions from business travel or employee commutes.
More information detailing the spoiler can be found in the following illustration:
The left side outlines the flow that we in EDA support to build chips and systems. Just as the three scopes of emission, the electronics impact has several scopes as well. First, there is the actual consumption of IT devices and infrastructure as outlined above. As a development ecosystem, we help with reducing direct power consumption with tool advancements during design and development, innovative new architectures, and progress in semiconductor and manufacturing technologies. Going back to the Foley quote from above, the electronics we enable and eventually helps build with a lot of the contributors to emissions. For instance, AI/ML processing that is electronics enabled can help with optimizing food usage and planning. Electronics sensors help with optimized agriculture production. Electronics are critical to electrification in transportation, better planning, and optimized shapes for the best energy consumption. And electronic sensors combined with AI/ML-enabled computing help with the automation of buildings and smart city infrastructure.
As shown, the energy impact of electronics has multiple layers and has to be considered holistically. More electronics devices and sensors consume more energy but enable new behaviors and processes that offset the direct consumption.
The third layer is what end consumers can do at a personal level. As an electronics industry, we enable insatiable consumer demand. This consumer adoption is driven by confidence, convenience, and collaboration, as we showed last year in the report “Hyperconnectvity and You – A Roadmap for the User Experience.”
The electronics industry clearly has a critical responsibility. You can find some interesting background reading from Bentley’s Rodrigo Fernandes in “Can Engineers Save the Planet from Climate Change and its Effects?” Looking at all this, there is no one company that can make an impact all by itself. I recently brought together a mix of EDA and IP providers, semiconductors, systems, and software developers for a panel “What can EDA and the Electronics Ecosystem Do for Greener Electronics to Save the Planet?” It was a fascinating conversation.
I am an optimist. We have the brightest minds in the ecosystem focusing attention on this, and the pace of innovation in this space makes me hopeful.
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
That may be true of the more mainstream data centers, BUT when you take a look at the extreme energy requirements of crypto-mines, that tells a very different story. There’s a reason why Texas asked all their crypto-mine facilities to shut down in the midst of an epic heat wave and demand for air conditioning recently.