Computer Architecture Today

Informing the broad computing community about current activities, advances and future directions in computer architecture.

We need to act on climate change now. With global emissions continuing to rise at an alarming rate there is an urgent need to adopt measures to reduce emissions. The UN calls climate change “the defining issue of our times”. It is imperative that we, the computer architecture community, do our part in this effort.

Our conferences have extremely large carbon footprints on a per-person basis due to the amount of air travel involved. It is considered that “Euro for Euro, hour for hour, flying is the quickest and cheapest way to warm the planet”. For example, an average global citizen produces about 5 metric tons of emissions per year. A person flying economy class from New York City to London and back causes about 1.67 metric tons of emissions or roughly one third of the emissions an average person produces in a year. Given these statistics, we need to carefully consider how to reduce the air travel incurred due to our conferences.

Here are some ideas that I have come across (through formal and informal channels) that could help us reduce our carbon footprint. Similar ideas have appeared in other forums. In fact, the SIGPLAN Climate Committee has compiled a detailed report on different ways in which we can make our conferences climate friendly. 

Making virtual participation a first-class option: With the proliferation of high-quality video conferencing/streaming and communication tools, an enriching virtual conference experience (for both authors and non-authors) has never been more feasible. There are several different ways in which virtual participation can be arranged and we will need to experiment with them to arrive at the right one for our community. For example, see Natalie Enright Jerger’s discussion of alternatives for in-person PC meetings.

A first-class virtual participation option reduces the need to travel to the conference and also has the benefit of improving accessibility to members of our community who cannot make it to the conference. However, it will also reduce the opportunities to build personal relationships and gain visibility in the community. With careful experimentation, I hope we can strike the right balance. ASPLOS 2020 moving to an asynchronous virtual conference model (amid trying circumstances) has a silver lining that it helps us learn more about the pros and cons of virtual participation. 

Conference co-location: Every year, the architecture community holds four major conferences (ISCA, MICRO, HPCA, and ASPLOS) and several other smaller and more focused conferences. We should consider co-locating our conferences, even major ones. Larger co-located conferences, perhaps organized over more days or with more parallel tracks will incur less air travel than our current model of having multiple, smaller conferences throughout the year. While larger co-located conferences reduce air travel, they bring their own sets of challenges. They incur higher organization overhead, they could be overwhelming, especially for new researchers, etc. We need to strike the right balance. Moreover, event co-location doesn’t only have to be at the conference level. The ISCA 2020 PC meeting was explicitly co-located with HPCA 2020 to reduce air travel involved. Such efforts must be recognized and encouraged by the community.

Optimize conference location: For every conference, there are usually multiple locations that are considered before settling on one. Apart from the usual set of considerations (for example, cost and location diversity), we should also consider the carbon footprint of traveling to the location. Well connected locations with direct flights from around the world should be preferred as direct, long flights are much more climate friendly (on a per-mile basis) than short connecting flights. ACM already provides a tool to help calculate the carbon footprint of air travel incurred due to conference location. The goal is not to simply limit our conferences to being held in a few major cities all the time, but, instead we should be factoring in the climate costs along with other usual (and important) considerations.

Carbon offsetting: Carbon offset schemes allow us to compensate for air travel by promoting activities that reduce global emissions like investments in clean energy installations and reforestation drives. For example, as per Delta, the emissions caused by the afore-mentioned single, economy class round trip from New York City to London can be offset by a donation of USD 6.64 to the Guatemalan Conservation Coast Project. The cost of these offsets pales in comparison to the typical costs of attending a conference (usually about USD 1500 in registration, hotel, and air travel costs). In fact, ACM already has a carbon offset program mechanism that conferences can adopt. As of now, this program allows attendees to calculate the appropriate offset amount and make a personal donation. While personal donations are great, one can also imagine arrangements where the conference acquires sponsorship to buy offsets for the whole conference.

Adopting carbon offsetting in our conferences seems like the obvious first step we should be taking as a community. However, carbon offsetting is no panacea. The efficacy of offsetting techniques are not clear, especially in the long term. It is recommended that carbon offsetting be treated as a measure of last resort, after all other options to reduce emissions have been exhausted. And, it is also necessary to find the right programs to donate to due to concerns over fraud and effectiveness. 

Acting on climate change now is important. There are several ideas we could pursue to reduce the carbon footprint of our conferences, each with its own pros and cons. The SIGARCH leadership must consider forming its own committee on climate to solicit and evaluate different ideas to make our conferences more climate friendly. The time to act was yesterday.

About the Author: Aasheesh Kolli is an Assistant Professor at Penn State University and focuses on developing next-generation memory systems.

Disclaimer: These posts are written by individual contributors to share their thoughts on the Computer Architecture Today blog for the benefit of the community. Any views or opinions represented in this blog are personal, belong solely to the blog author and do not represent those of ACM SIGARCH or its parent organization, ACM.