Workshop

Type of workshop

• Half day
• Expected number of participants: 15-25

Scope and objectives

We want to bring a group of like-minded and critical colleagues together to chart our understanding of the immediate materiality of semiconductor industry and its far-reaching consequences. We will investigate the mature node microchips and their abundance from an ecological standpoint.

We will first look into the production side of the semiconductor industry. We will try to identify the global volume of microchips being manufactured and the complex supply chain behind their distribution7. Just packaging and assembly on a circuit board can already result in a microchip travelling between multiple countries many times over. Our analysis will include not only the ‘standard’ microchips but also the OSD (optical, sensor, and discrete) components, which make up 2/3 of the semiconductor component market.

As a faulty/broken microchip itself can never be fixed (very rarely with extremely expensive machines), anything containing a microchip can only be mended by replacing it. Furthermore, on a circuit board with multiple microchips, it is often impossible to determine which one is broken and how to replace it; the whole board, along with all other components, is replaced. We will try to determine which everyday devices use the most microchips by volume and the implications for eWaste generation.

We will then analyse the enabling impact of the microchip industry from a socio-technical perspective. This is similar to the Scope-3 downstream emission of an industry but beyond just the greenhouse gas. The enabling impact will try to answer how the abundance of cheap electronics has changed the way ICT devices are consumed. The lack of circularity and the unintentional obsolescence that result from this massive, never-ending innovation and production cycle. Our goal is to produce a written briefing document that sets out the challenges and importance of this domain for the ICT4S community and for engaging with policy and strategic leaders on this topic.

Prerequisites

The workshop participants will be asked to submit short abstracts on topics broadly connects the list below.

• The global volume of microchip production and their final destination
• Supply chain of these vast number of cheap components and hidden GHG emission
• Environmental impact and enabling effect of these invisible and indispensable components
• Societal perspectives and governance of digital products, supply chains, circularity, and e-waste.

Schedule

The participants will be asked to do a short presentation (20min) on their abstract, followed by a 10 min discussion. We do not expect more than 3 presentations. This time will be used to create individual research project on more specific topics. This will be done by multiple round table discussions continuing the topics/issues highlighted in the morning session.

Background

This workshop is a continuation of the TCICT (True Cost of ICT) series we started in 2023. The original workshop (TCICT: From Materiality to Techno-Solutionism) looked into the materiality of ICT, primarily from the requirement of semiconductor/microchip industry. It also aimed to address the narrative of ‘techno-solutionism’ that arises from the continuous innovation in the microchip sector (aka, Moore’s Law).

The ICT industry is essential to a vast array of gadgets, products and services that proliferate in numbers beyond most of our imaginations. However, it is the semiconductor microchips, one of most important components in this digital world, that glues these technologically enabled industries together. The growth of microchips is unprecedented compared to any other industrial sector. Following Moore’s law, there has been an efficiency increase (i.e., transistors per chip) by a million times in 50 years. All of the latest AI advances are due to the most cutting-edge microchips designed by Nvidia (and others) and fabricated at TSMC, Taiwan. These enormously complex objects consist of over 200 billion transistors (each around 3nm in size!). There is no doubt that the energy and material cost of such an incredible feat of engineering is far beyond anything ever produced.

The ‘AI hype’ has rightly focused our attention on the environmental costs of producing such technology (the ‘advanced node’ transistors) and also that of their subsequent uses in ever more gigantic data centers. As microchip manufacturers keep pushing further in the name of ‘efficiency’ (the industry has plans for sub-nm sized transistors in the next 5 years), the true ecological impact behind them will need to be scrutinised even more closely. While future generations of these devices are likely to need particle accelerators for their production, not all microchips come remotely close to such complexity. There are, in fact, a vast world of microchips that originally produced over 30 years ago (many lifetimes in tech world) are still in use. These older generations (often called ‘mature nodes’ or ‘lagging nodes’) of microchips remain in great demand and are often produced in the same volume as they were originally introduced. Yet, often neglected from scrutiny.

This persistence of very old technology is not a random quirk of the industry, but an important feature. Since several billion dollars are necessary to develop a ‘node’ to maturity, manufacturers keep using them (often to produce very cheap microchips) for as long as possible. A 350nm microchip, first released in 1995, is not only 100 times less ‘efficient’ but at least 12 generations old! With most older generations still active, the total number of microchips produced per year keeps accumulating as new ones get added.

Ultra-cheap microchips are abundant and create the vast bulk of the >400 billion pieces sold per year. While the latest Nvidia GPU might cost a few million and are nearly impossible to buy on the open market, the average cost of a microchip is <$1 and most are readily available. This abundance comes with a huge environmental cost. The microchips do not exist in vacuum, they are packaged into every possible gadget around us, whether necessary or not. They make everyday objects ‘smart’, but this also increases their obsolescence. From disposable vapes to smart toys, cheap microchips enable the production of billions of new objects that will be discarded very soon. For example, over one billion ‘fast-tech’ gadgets (e.g., handheld fans) were bought by UK consumers alone in 2024, half of which have already been disposed of. This is nothing short of an environmental crisis, something made possible by the abundance of cheap digital components whose true costs are hidden. The expensive advanced nodes of today will become mature nodes in a few years and be available at low cost to flood the consumer electronics market further.

References

1. Miller, C. Chip War: The Fight for the World’s Most Critical Technology. (Simon & Schuster UK, London, 2022).
2. McGregor, J. The True Nature Of Moore’s Law – Driving Innovation For The Next 50 Years. Forbes (2022).
3. Roussilhe, G., Pirson, T., Bol, D. & Mitra, S. Purer than pure: how purity reshapes the upstream materiality of the semiconductor industry. Preprint at arXiv (2025).
4. IEEE IRDS. IEEE International Roadmap for Devices and Systems - IEEE IRDSTM. https://irds.ieee.org/ (2023).
5. Boyd, J. Is the Future of Moore’s Law in Particle Accelerator? IEEE Spectrum vol. June (2024).
6. Banfield-Nwachi, M. It’s cheap but it’s not disposable’: why fast tech is a growing waste problem. The Guardian (2025).
7. Ou, S., Yang, Q. & Liu, J. The global production pattern of the semiconductor industry: an empirical research based on trade network. Humanit Soc Sci Commun 11, 750 (2024).