How we're engaging the semiconductor industry to accelerate decarbonization

Whether you're streaming your favorite show, mapping driving directions home, or just sending an email, your day-to-day life depends on semiconductors. These chips are critical components of virtually every electronic device and serve as the backbone for AI, which we are utilizing to help solve some of the world’s most complex challenges, such as enabling hyper-accurate weather forecasting and driving new breakthroughs in drug discovery.

At the same time, semiconductor manufacturing has a significant environmental footprint. The industry is expected to emit 277 million metric tons of CO₂ equivalent by 2030, a figure comparable to the emissions of 10 cities the size of London.^{1, 2} The latest semiconductor technologies are twice as carbon intensive as earlier generations, creating a clear opportunity for industry intervention and innovation to lower their impact.³

Uniting industry leaders to respond to industry challenges

To help drive action, Google partnered with SEMI’s Semiconductor Climate Consortium to host the SEMI Global Executive Summit (GES) in Tokyo, in December 2025. We brought together more than 160 senior executives from nearly 100 companies across the entire semiconductor value chain — including chip manufacturers, equipment makers, and raw material suppliers — to establish a practical framework for emissions reduction. Together, we identified four areas where we can collaborate across the industry to have the biggest impact.

1. Reducing process gas releases into the atmosphere

Building the microscopic intelligence in chips requires using some of the most potent greenhouse gases (GHGs) ever identified, making them responsible for nearly all direct (or non-electricity) emissions emitted by a semiconductor fabrication plant (fab). For instance, sulfur hexafluoride (SF6) is the most potent gas identified by the Intergovernmental Panel on Climate Change (IPCC), with a Global Warming Potential (GWP) 24,300 times that of carbon dioxide.⁴ Because SF6 also stays in the atmosphere for 3,200 years, the gases used for one generation of electronics will impact the climate for the next 100 generations.⁵

The mitigation approach with the most direct impact is abatement, which involves destroying unconsumed process gases to convert them into less harmful substances before they are released into the atmosphere. Scaling this solution requires industry-wide collaboration to 1) improve how process gases are measured and neutralized, 2) advance abatement technologies (like with catalysts which allow for improved destruction with less input energy), and 3) share the data necessary to accelerate adoption across the supply chain. Sharing data reduces the risk of adopting a solution and moves low-carbon technologies from the lab to the fab more quickly.

2. Identifying alternative, lower-potency process gases

A long-term, high-impact opportunity involves developing new, alternative gases with considerably lower GWPs to replace the current highly potent process gases. This requires accelerated, collaborative industry research to identify and validate alternatives. Chip manufacturers, equipment makers who make the tools that consume these gases, and gas vendors are already working together in industry consortia to validate the performance of new alternatives. These alternative gases must still meet rigorous performance requirements while achieving sufficient adoption to justify their commercial development and economy of scale. Given the years-long horizon for widespread fab adoption, it is essential that this work begins without delay.

3. Increasing clean electricity in fabs

Maximizing the use of clean electricity is the most impactful lever for semiconductor carbon reduction. Fabs consume enormous amounts of energy (up to 100 MWh per hour), and over 80% of production capacity is in Asia, where energy grids often rely heavily on fossil fuels.^{6, 7} With more than 80 new fabs expected to begin production between 2025 and 2030, access to clean electricity faces constraints due to infrastructure, high costs for renewables, and limited purchasing options.⁸ Ensuring progress in key production regions like South Korea, Japan, and Taiwan requires alignment on policy agendas, joint advocacy, and even region-specific pilots that can test new procurement models.

4. Engaging and empowering suppliers to drive their own reductions

Although most environmental impact occurs at the fabs, the production of input materials (like silicon wafers, gases, and chemicals) is highly energy-intensive. There is an opportunity to empower all suppliers, not just fabs, to make emission-reduction progress. Ecosystem players must engage their own suppliers to establish shared reduction goals and support participation in regionally-focused clean electricity training programs like the Clean Energy Buyers Association’s Clean Energy Procurement Academy (co-founded by Google), which helps suppliers build the technical capabilities needed to drive progress.

Scaling industry-wide decarbonization through partnership

This unprecedented gathering of senior leaders across the ecosystem, extending beyond sustainability professionals, signals the shared commitment required to tackle this complex, interdependent problem. We are encouraged by the openness of leading manufacturers to share what is working and identify where partnership is needed. Pre-competitive knowledge sharing is essential to advance innovations — like gas substitution and abatement — from theoretical research to scalable, fab-ready solutions and to increase clean electricity access for fabs and their suppliers.

Moving forward, the entire industry must collaborate to reduce process emissions and increase clean electricity access in tangible, actionable ways. By aligning on technical ambitions, advancing collaborative R&D, and bringing a common voice to policy engagements, we can ensure the intelligence powering tomorrow’s AI and all other electronics is built on a sustainable foundation.