Tag: energy efficiency

  • Murakkab: New System Cuts AI Agent Energy Use and Cost by Over 70%

    Murakkab: New System Cuts AI Agent Energy Use and Cost by Over 70%

    Agentic workflows — AI-powered software systems that chain multiple models and tools to complete complex tasks — are becoming the backbone of cloud computing. But their fragmented design often wastes computation, energy, and money. Researchers from MIT and Microsoft have developed a new system called Murakkab that streamlines the design and deployment of these workflows, automatically optimizing them for speed, energy efficiency, and cost.

    With Murakkab, developers describe their application’s goal in plain language, and the system automatically selects the best AI models, tools, hardware configurations, and resource allocations. It adjusts these on the fly based on user priorities, such as minimizing costs or maximizing speed. In tests, Murakkab used only about 35% of the computation, 27% of the energy, and under 25% of the cost compared to traditional methods — without sacrificing performance.

    “Agentic workflows are getting very complicated and quickly becoming the backbone of what cloud providers are doing,” says Gohar Chaudhry, an MIT EECS graduate student and lead author of the paper presented at USENIX OSDI. “Energy usage is a huge concern, so we need to be very careful about how efficient these workflows are.”

    Murakkab also adapts dynamically when new models or hardware become available, eliminating the need for developers to manually reconfigure their systems. The researchers plan to expand the system to more complex workflows and larger computing clusters.

  • MIT Researchers Unveil Ways to Cut Data Center Energy Use and Boost Sustainability

    MIT Researchers Unveil Ways to Cut Data Center Energy Use and Boost Sustainability

    A new study from MIT suggests that flexibility in the timing of electricity consumption at data centers could lower consumer costs. The research highlights how adjusting when energy is used can help manage demand and reduce strain on the grid.

    In related work, MIT researchers have developed a system called Murakkab that improves the speed and energy-efficiency of AI agents. This innovation optimizes the design and deployment of multistep workflows powering AI applications.

    Another project introduces a computer model that enables bridges and buildings to use less material while remaining buildable. The approach bridges the gap between optimized design and real-world construction constraints.

    MIT Professor Susan Solomon was named a 2026 Tang Prize laureate for her groundbreaking work on atmospheric chemistry, which helped lay the foundation for ozone layer recovery and demonstrated the lasting impacts of carbon emissions on climate.

    The MIT Environmental Solutions Journalism Fellowship has expanded climate reporting through local messengers, reaching nearly 3 million readers and listeners with community-centered coverage.

    A startup co-founded by two MIT researchers, Ferveret, uses a nuclear-inspired cooling system to reduce energy and water needed for cooling chips that power AI, making data centers more sustainable.

    Other MIT projects explore low-cost personal cooling and emissions-free air conditioning to address extreme heat, while researchers develop innovative carbon capture methods and a low-cost technique to extract lithium from rocks.

    A study on wetlands preservation shows that tradeoffs between conservation and development can be less stark with a policy featuring tradeable offsets and taxes. The MIT Asia Real Estate Initiative expands into booming Asian cities, and MIT Sloan fellows share insights on leading a sustainable future.

    For most U.S. drivers, electric vehicles offer emissions benefits and cost savings, with individual driving patterns and regional electricity mix playing key roles.