CCC’s most recent Community Chat, Powering the Solution: Balancing Computing’s Energy Demands with its Contributions to Sustainability, took place on October 8, 2026. Moderated by CCC Council Executive Committee Member Michela Taufer (University of Tennessee, Knoxville), the webinar also featured fellow authors of the recent Communications of the ACM (CACM) opinion article Balancing Sustainable Computing and Computing for Sustainability.

Watch the Recording Here

Lightning talks from Chandra Krintz (University of California, Santa Barbara), Ana Radovanovic (Google), and Adam Wierman (California Institute of Technology) were followed by a Q&A, joined by Elizabeth Bradley (University of Colorado Boulder), Siddharth Parida (Embry-Riddle Aeronautical University), Masood Parvania (University of Utah), Eve Schooler (University of Oxford), and Bronis de Supinski (Lawrence Livermore National Laboratory).

Two Sides of the Energy Demand Problem

Artificial intelligence (AI) and high performance computing (HPC) are driving advances in healthcare, weather forecasting, and emergency response, but their energy demand has already outpaced what power grids can supply. The International Energy Agency projects that global data center electricity use will more than double between 2024 and 2030. 

At the same time, these technologies play a direct role in addressing environmental problems, supporting work on sustainable agriculture, more efficient batteries, recyclable renewable energy technologies, and sustainable plastics. The authors argue that sustainable computing (reducing computing’s footprint) and computing for sustainability (using computing to address environmental challenges) must be pursued together, producing a multiplier effect that neither achieves alone.

Michela Taufer stressed that efficiency alone is not enough. Individual computations can use less energy even as total consumption rises, because the volume of computing keeps growing. The article calls for research, technology development, and governance to advance together, drawing on computer science, environmental science, power systems, social science, policy, and industry, and closes with open questions for the community.

Three Perspectives on Balancing Sustainability

Which computations are worth doing. Drawing on agriculture and ecology, Chandra Krintz explained that in physical settings, timing matters as much as accuracy: a prediction that arrives after an irrigation or pest management decision has little value. Her team predicts the value of an HPC job before running it. Across roughly 100 jobs on NERSC systems, this improved accuracy gained per core hour by 35% compared with running every job, while using 90% fewer core hours. “The principle is, compute when computation matters,” Krintz said.

When and where to run it. Ana Radovanovic described how data center growth is straining grid capacity and delaying interconnection, making it essential for data centers to be good grid citizens. “Data center fleet growth today cannot happen without a sustainability-aware design throughout its lifecycle, from siting to workload scheduling,” she said, then explaining how Google’s carbon-aware computing shifts flexible workloads in time and location toward cleaner power. Newer approaches include curtailing machine learning training during grid emergencies, partnering with virtual power plant operators, and siting near clean generation.

What it means for host communities. Adam Wierman noted that a majority of Americans now oppose building a data center in their area. “Communities aren’t worried about the long-term trade-offs, they’re worried about what they are experiencing today,” he said. Those costs include noise, rising electricity rates, strain on water systems, and pollution from on-site generators, and they fall disproportionately on low-income communities. Well-designed partnerships can also bring revenue and infrastructure investment, and Wierman sees research as a way to help communities weigh those trade-offs.

From the Broader Discussion

The webinar closed with a Q&A, allowing attendees to ask the questions they cared the most about. This led to a few key discussions related to the broader sustainability question.

What to optimize. Asked what to optimize for if efficiency only drives more consumption, Liz Bradley pointed to the need to assess true cost over the full life cycle, including impacts on disadvantaged communities. That cost is hard to pin down, she noted, because many impacts are easy to overlook and policymakers and investors tend to prioritize the short term, making the right metric “a really hard socio-technical problem.” Radovanovic added that this will take several metrics, including embodied emissions, water use, and how much load can shift when the grid needs it. Krintz also noted that resource data rarely reaches the people deciding what to compute.

Policy and education. Wierman encouraged researchers to show what is possible so policymakers can develop technologically informed policy. Schooler called for teaching metrics beyond performance, from energy and carbon to water and noise, early in computing education.

Standards and disclosure. Asked about report cards for data centers, Schooler pointed to standards work at the Open Compute Project, IEEE, IETF, and ISO. Because standards take time, she suggested pilot projects with community and industry partners as a faster way to put shared metrics into practice. Wierman added that communities need metrics and public disclosure urgently in order to evaluate proposed data centers.

Advice for early-career researchers. Parvania encouraged working on practical problems with industry partners, Krintz urged building interdisciplinary networks early, and Schooler pointed to open questions in the economics of resource use and more decentralized infrastructure.

Watch the Webinar and Read (and Cite!) the Article

If you missed the event, you can watch the recording and read the CACM article below:

Powering the Solution: Balancing Computing’s Energy Demands with Its Contributions to Sustainability Webinar

“Balancing Sustainable Computing and Computing for Sustainability” CACM Article

We hope you will share both resources with colleagues and students working on computing and sustainability, as well as cite them if they prove useful for your own research. For those interested in the topic, we also encourage you to subscribe to CCC’s LinkedIn page or CCC Blog at the links below to be notified about a forthcoming report from the Full-Stack Sustainability workshop in February 2027.

Tune in to the CCC LinkedIn Showcase Page for updates and more Community Chats like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

This material is based upon work supported by the U.S. National Science Foundation (NSF) under Award Nos. 2300842 and 2619366. These awards support the Computing Community Consortium (CCC), a programmatic committee of the Computing Research Association (CRA). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.