Introduction:

The current landscape of computer science is rich with innovation. Powerful new technologies like large language models, generative AI programs, cloud computing options, improvements in computer vision, novel robotic designs, and more are actively impacting how businesses approach their technical futures. To satisfy the expectations of modern consumers and stay competitive in the market, companies strive to make their systems as intelligent and secure as possible. 

Unfortunately, the enormous gains promised by advancements such as machine learning come with an equally enormous cost, both financial and environmental. By some estimates, ChatGPT costs over $700,000 to run per day. Training a medium-sized language model is known to release 626,000 pounds of carbon dioxide into the atmosphere, which is equivalent to the lifetime emissions of five cars. And the ever-increasing volume of cloud data storage—an estimated 180 billion terabytes of data will be created in 2025—bring with it huge sustainability concerns, as storing just a single terabyte of this data can produce up to 2 tons of carbon dioxide annually. 

As customers grow more environmentally conscious and governments begin to put pressure on companies to make sustainable choices, these challenges could prove significant to a company's bottom line. Aside from the energy cost, the raw price tag on some of these game-changing inventions makes them risky to implement, difficult for smaller businesses to leverage, and may strain what should be a dynamic and nimble field of discovery. In order to reap the benefits of these technologies, we need sustainable solutions that can both fuel this boom of innovation and prevent environmental damage and spiraling costs. 

To address this challenge, MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) is creating a new research initiative that will focus on sustainable computing in partnership with industry. Next Generation Software Efficiency (NGSE) will be focused on advancing the future of efficient software options and facilitating the transition of fundamental academic knowledge to real-world technological solutions. We are inviting a small group of organizations to help us build this research collaboration and usher in a new generation of efficient, effective, and environmentally friendly computation. 

Technical Research Aims:  

·      The development of more efficient algorithms: As Moore’s Law comes to an end, designing efficient algorithms that make full use of the available computation becomes an increasingly vibrant area of study. Many of today’s legacy systems depend on algorithms and computer languages that were developed in the days before such limitations and therefore leave potential speed and efficiency on the table. By optimizing computational processes, efficient algorithms reduce processing times, conserve valuable resources, and minimize energy consumption, leading to significant cost savings and a reduced environmental footprint. Moreover, in the context of big data and artificial intelligence, efficient algorithms enable faster data analysis and more accurate predictions, driving innovation and advancements across various domains.

·      Creating systems and architecture that maximize available computing power: Also relating to the end of end of Moore’s Law, the architecture that current legacy systems depend on is often based on designs that did not prioritize low energy consumption and making the most effective use of resources. By leveraging concepts like parallel processing, load balancing, and distributed computing, developers can fully harness the power of current processors and GPUs.

·      Rethinking the traditional way cloud computations operate: Cloud computing has been an enormous boon to the economy, but the pressure on companies to have data immediately accessible via cloud is both expensive and currently unsustainable. To support the convenience and innovation of cloud computing while addressing its setbacks, researchers are investigating key topics like applying machine learning to foresee and prevent wasteful cloud outages and more effectively manage limited resources.

·      Adapting current models for alternative uses: Knowing the monetary cost and energy footprint of programs such as large language models (LLMs), a natural source of inquiry is whether or not previously trained models can be adapted for purposes they were not initially intended or designed for. Early research in this area shows promise in, for example, few- or zero-shot training of LLMs in specialized domains, offering an alternative to companies looking to fill an ML need without the price tag of creating a large model for that exclusive purpose.

·      Exploring new, compact, and sustainable models: Another option being explored in ML research is the creation of smaller, more targeted, and less expensive models. Rather than training an enormous LLM like ChatGPT, some situations can be better served by compact versions trained on very specific, curated data, making them perfectly targeted for a given use, easier to create, and more sustainable to run than their large counterparts.

·      CPU efficiency and re-envisioning data centers: Datacenters are the foundation of many areas of computer science, but they are energy-hungry, inefficiently designed, and current operating systems often struggle to keep up with increasing demand. Recent work redesigning datacenters could help with methods such as proactive program scheduling, resource management, and better storage hardware.

NGSE Goals: 

• To connect the research in the lab to real-world challenges for which there is no current commercial solution and to explore the future impact of sustainable computing. 

• To provide visibility and bring together MIT CSAIL researchers working across disciplines on a variety of sustainable computing challenges.

• To seed projects that cut across multiple areas addressing sustainable computing: CPU efficiency in datacenters, efficiency in computer security, efficient training of machine learning models on edge devices, deep learning models, representation learning, generative modeling, optimally efficient algorithms, using pre-existing LLMs in alternative applications, simulation-based training and more.

• Working jointly with industrial partners, to create a roadmap of sustainable computing transformative technologies that will enable a positive impact in this space.

Leadership & Governance: 

NGSE@CSAIL will be led by MIT Professor Adam Belay, whose research focuses on operating systems, runtime systems, and distributed systems. Professor Belay works to build systems that make datacenters faster, more efficient, and more secure.

Administratively, NGSE@CSAIL will be run by CSAIL Alliances and will interface with research groups across CSAIL to facilitate research projects poised to generate long-term impact in the world in the space of sustainable computing.

Industry members of NGCE@CSAIL will each have one seat on the Executive Board along with the faculty director and the Executive Director from CSAIL Alliances. The board will meet at various times throughout the year to discuss research and challenges faced by members. Problem statements will be circulated broadly across the lab to all research disciplines. 

There will be an annual meeting on the projects undertaken in this initiative for initiative members and interested community members. NGCE@CSAIL will also have a session at the CSAIL Alliances Annual Meeting. 

There will also be a public event to share the work being done in the space and provide opportunity to dialogue with additional organizations.