Home > News > All News > Mosharaf Chowdhury

Mosharaf Chowdhury

Up to 30% of the power used to train AI is wasted. Here’s how to fix it.

Smarter use of processor speeds saves energy without compromising training speed and performance.

Four papers by CSE researchers at OSDI 2024

New papers by CSE researchers cover topics related to the design and implementation of systems software.

Poster session showcases student-developed GenAI software systems

Prof. Mosharaf Chowdhury’s Systems for GenAI course closed with a poster session highlighting student projects.

CSE researchers receive Mozilla funding for research on AI energy use

The researchers were selected as recipients of the 2024 Mozilla Technology Fund for Zeus, an effort to measure and optimize the energy consumption of machine learning.

Open-source training framework increases the speed of large language model pre-training when failures arise

Pipeline templates strike a balance between speed and effectiveness in resilient distributed computing.

Fan Lai awarded Kuck Dissertation Prize for thesis on minimalist systems for machine learning

The annual award recognizes the most impactful dissertations by PhD researchers in CSE.

CSE researchers present new findings at OOPSLA and SOSP

Several researchers in CSE are presenting papers at the two conferences on programming languages, operating systems, and more.

Congrats to CSE alums who have accepted faculty positions

Congrats to these new faculty!

Power-hungry AI: Researchers evaluate energy consumption across models

A new tool designed by researchers at the University of Michigan allows users to compare the energy efficiency of AI-powered language models.

New technique for memory page placement integrated into Linux kernel

A novel mechanism designed by CSE researchers that automatically tiers memory pages has been deployed in the Linux operating system.

Mosharaf Chowdhury receives Google Research Scholar award for research on resilient deep learning

Chowdhury is working to develop new fault-tolerant techniques to enable deep neural networks to continue training even when failures occur

Two CSE PhD students named Machine Learning and Systems Rising Stars

Fan Lai and Jiachen Liu have been selected to join a competitive cohort of early-career researchers looking to promote progress, collaboration, and research excellence in machine learning and systems.

Optimization could cut the carbon footprint of AI training by up to 75%

Deep learning models that power giants like TikTok and Amazon, as well as tools like ChatGPT, could save energy without new hardware or infrastructure.

Fan Lai earns Towner Prize for Outstanding PhD Research

The award recognizes creative and outstanding research achievements.

Researchers cut down on AI's carbon footprint with new optimization framework

Zeus automatically adapts the power usage of deep learning models to chase clean electricity sources throughout the day

Open source platform enables research on privacy-preserving machine learning

Virtual assortment of user devices provides a realistic training environment for distributed machine learning, protects privacy by learning where data lives.

Multi-institute project "Treehouse" aims to enable sustainable cloud computing

"We are buying thousands of GPUs and running them at full speed, and no one really knows just how much energy is being spent in the process."

Enabling efficient, globally distributed machine learning

A group of researchers at U-M is working on the full big data stack for training machine learning models on millions of devices worldwide.

Four papers with Michigan authors at SIGCOMM 2021

ACM SIGCOMM's annual conference is the leading conference in data communications and networking in the world.

Human resilience study to benefit from new data privacy technique

Prof. Mosharaf Chowdhury is leading development of a new machine learning application that will protect the privacy of participants.

Mosharaf Chowdhury named Morris Wellman Professor

Chowdhury is an expert in network-informed data systems design for big data and AI applications.

“Hiding” network latency for fast memory in data centers

A new system called Leap earned a Best Paper award at USENIX ATC ‘20 for producing remote memory access speed on par with local machines over data center networks.

Enabling fairer data clusters for machine learning

Their findings reduce average job completion time by up to 95% when the system load is high, while treating every job fairly.

Big data, small footprint

How changing the rules of computing could lighten Big Data’s impact on the internet.

Five papers by CSE researchers presented at NSDI

The teams designed systems for faster and more efficient distributed and large-scale computing.

Chowdhury receives VMWare Award to further research on cluster-wide memory efficiency

Chowdhury’s work has produced important results that can make memory in data centers both cheaper and more efficient.

Chowdhury wins NSF CAREER award for making memory cheaper, more efficient in big data centers

Chowdhury connects all unused memory in a data cluster and treats it as a single unit.

Two solutions for GPU efficiency can boost AI performance

Chowdhury’s lab multiplied the number of jobs a GPU cluster can finish in a set amount of time

Designing a flexible future for massive data centers

A new approach recreates the power of a large server by linking up and pooling the resources of smaller computers with fast networking technology.

A breakthrough for large scale computing

New software finally makes ‘memory disaggregation’ practical.

Jack Kosaian selected for NSF Graduate Research Fellowship

Jack has enjoyed involvement in research across diverse domains within the College of Engineering.

Mosharaf Chowdhury receives ACM SIGCOMM Dissertation Award

Prof. Chowdhury bridges the gap between application-level performance and network-level optimizations through the coflow abstraction.

Mosharaf Chowdhury receives Google Faculty Research Award

The project aims to create a new software stack for analytics over geo-distributed datasets.

Eleven New Faculty Join CSE

We're building a bigger, better CSE.