NextGArch Lab

NextGArch Lab
Pioneering the Future of NextGen Systems and Networks!
At the University of Michigan, NextGArch Lab is dedicated to revolutionizing computer systems, networks, and architectures through cutting-edge research and innovation in CSE and EECS.
Learn More
About NextGArch Lab
Our Mission
At the NextGArch Lab, we conduct innovative research in computer systems, networks, and architectures. Our work involves developing domain-specific abstractions, compilers, and architectures for networks and systems, with applications in AI/ML, self-driving networks, cloud/edge computing, and 5G/6G.
Research Areas
Domain-Specific Systems
New programming models, runtime systems, and architectures for nextgen high-performance and scalable computing.
Domain-Specific Networks
Novel networking protocols, architectures, and algorithms for efficient and reliable data communication.
Domain-Specific Architectures
Innovative architectures for future computing systems, including hw/sw co-design and accelerators for line-rate and proficient ML/AI.
Systems+X and Networks+X
Leveraging cross-domain insights—e.g.,  X = ML/AI—to push the boundaries of efficiency, scalability, and innovation in distributed and networked systems.
Latest News
May 29, 2025
Prof. Shahbaz gave an invited talk during the IA382: Seminar in Computer Engineering at FEEC/UNICAMP!
feec-seminar-comp-eng.github.io
Fast, Flexible, and Intelligent Next-Generation Networks and Systems
Abstract # Maintaining strict security and performance objectives in next-generation cloud and edge networks demands that compute-intensive management and control decisions are made on the current state of the entire network (e.
May 9, 2025
SpliDT accepted to SRC TECHCON '25. Many congratulations to Marilyn Rego (Undergraduate Lead), Murayyiam Parvez, Annus Zulfiqar, and the team!
www.src.org
TECHCON 2025 (Event E007206) - SRC
Join us for TECHCON 2025, the premier conference for innovation and excellence in the semiconductor industry! Taking place September 7-10, 2025, at the Renaissance Hotel in Austin, Texas, this exclusive, members-only event brings together industry leaders, researchers, students, and recruiters to explore cutting-edge advancements, exchange ideas, and build the future of microelectronics.
May 8, 2025
Gigaflow and SpliDT accepted into this year’s Google Summer of Code (GSoC) cohort. Congratulations to Advay Singh and Sankalp Jha for their successful proposals and for leading the projects this summer!
linkedin
📢🎉 Thrilled to share that NextGArch Lab’s projects, Gigaflow and SpliDT,… | Muhammad Shahbaz
📢🎉 Thrilled to share that NextGArch Lab’s projects, Gigaflow and SpliDT, have been accepted into this year’s Google Summer of Code (GSoC) cohort! The projects will be led by Advay Singh, a rising undergraduate in Computer Science and Engineering at the University of Michigan, and Sankalp Jha from AJAY KUMAR GARG ENGINEERING COLLEGE, GHAZIABAD—supported by an incredible team of mentors from Google and NextGArch Lab!  NextGArch Lab: https://lnkd.in/geqB2Xyq Gigaflow: https://lnkd.in/gEvn5WE9  Go
Apr 23, 2025
Many congratulations to Annus Zulfiqar on being inducted into Sigma Xi, The Scientific Research Honor Society!
www.sigmaxi.org
Home
Top awards were given by judges and public vote for the 2025 virtual student competition.
Apr 22, 2025
Ertza Warraich inducted into the Phi Kappa Phi Society at Purdue University. Huge congratulations, Ertza!
honors.purdue.edu
Phi Kappa Phi - 
			 John Martinson Honors College -  Purdue University
“A global network of the best and brightest, a community of scholars and professionals building an enduring legacy for all generations”
Apr 19, 2025
Woot! Bilal Saleem is off to Arrcus this Summer!
Arrcus
Hyperscale, Multi-Cloud Networking Software
Bringing transformative 5G & multi-cloud networking to CSPs to lower 5G infrastructure cost, expedite service delivery and monetize cloud infrastructure faster. Test drive the ACE platform for free.
Apr 4, 2025
O'MINE accepted to Euro S&P '25. Congratulations, Enkeleda Bardhi, Ali Imran, and the team!
eurosp2025.ieee-security.org
IEEE European Symposium on Security and Privacy 2025 (EuroS&P)
Since 1980, the IEEE Symposium on Security and Privacy has been the premier forum for
                            presenting
                            developments in computer security and electronic privacy, and for bringing together
                            researchers and
                            practitioners in the field. Following this story of success, IEEE initiated the
                            European
                            Symposium on Security and Privacy
            
Mar 20, 2025
HardHarvest accepted to ISCA '25. Congratulations, Jovan Stojkovic, Chunao Liu, and the team!
linkedin
CPU cycles in shared datacenters are *incredibly* precious—we can&#39;t afford… | Muhammad Shahbaz
CPU cycles in shared datacenters are *incredibly* precious—we can&#39;t afford to let them sit idle. With HardHarvest, we offer a lightning-fast way to share these…
More …
Our Team
Muhammad Shahbaz
Principal Investigator (PI)
Assistant Professor, U-M (CSE)
Click here
Ertza Warraich
Ph.D. Student (Purdue)
Focus: ML Systems and Networks, Security, and AI
Click here
Annus Zulfiqar
Ph.D. Student (U-M)
Focus: HW/SW Co-Design, Virtual Networks, and In-Network AI
Click here
Bilal Saleem
Ph.D. Student (Purdue)
Focus: Cloud-Native Systems, Edge Computing, and 5G
Click here
Murayyiam Parvez
Ph.D. Student (Purdue)
Focus: Network Security, In-Network ML, and Programmable Data Planes
Click here
Ali Imran
Ph.D. Student (U-M)
Focus: Programmable Data Planes, In-Network ML, and Agentic Systems
Click here
Qizheng Zhang
Ph.D. Student (Stanford), co-advised with Kunle Olukotun
Focus: ML and Agentic Systems, In-Network ML, and Video Streaming
Click here
Omar Basit
Ph.D. Student (Purdue), co-advised with Y. Charlie Hu
Focus: AR/VR, 5G Systems and Cloud/Edge Computing
Click here
Marilyn Rego
B.S. Student (Purdue)
Focus: Agentic Systems, In-Network ML, and Domain-Specific LLMs
Click here
Venkat Kunaparaju
B.S./M.S. Student (Purdue)
Focus: HW/SW Co-Design, Virtual Networks, and Cloud Computing
Click here
Advay Singh
B.S. Student (U-M)
Focus: Cloud Computing, and ML Systems and Networks
Click here
More …
Selected Publications
TECHCON '25
SpliDT: Partitioned Decision Trees for Scalable Stateful ML Inference at Line Rate
Marilyn Rego, Murayyiam Parvez, Annus Zulfiqar, Roman Beltiukov, Shir Landau Feibish, Walter Willinger, Arpit Gupta, Muhammad Shahbaz
Euro S&P '25
O'MINE: A Novel Collaborative DDoS Detection Mechanism for Programmable Data-Planes
Enkeleda Bardhi, Chenxing Ji, Ali Imran, Muhammad Shahbaz, Riccardo Lazzeretti, Mauro Conti, Fernando Kuipers
ISCA '25
HardHarvest: Hardware-Supported Core Harvesting for Microservices
Jovan Stojkovic, Chunao Liu, Muhammad Shahbaz, Josep Torrellas
PDF
ASPLOS '25
Gigaflow: Pipeline-Aware Sub-Traversal Caching for Modern SmartNICs
Annus Zulfiqar, Ali Imran, Venkat Kunaparaju, Ben Pfaff, Gianni Antichi, Muhammad Shahbaz
PDF
Artifact
Computer Science and Engineering
Streamlining cloud traffic with a Gigaflow Cache
Gigaflow improves virtual switches for programmable SmartNICs, delivering a 51% higher hit rate and 90% lower misses.
Tech Xplore
Gigaflow cache streamlines cloud traffic, with 51% higher hit rate and 90% lower misses for programmable SmartNICs
A new way to temporarily store memory, Gigaflow, helps direct heavy traffic in cloud data centers caused by AI and machine learning workloads, according to a study led by University of Michigan researchers.
15:34
YouTube
Gigaflow - Pipeline-Aware Sub-Traversal Caching for Modern SmartNICs (ASPLOS 2025)
Learn about Gigaflow: a high hit rate, SmartNIC-native cache for virtual switches (like OVS) that expands rule space coverage by two orders of magnitude and reduces cache misses by up to 90%. 

This work was presented as ASPLOS'25.
NSDI '25
OptiReduce: Resilient and Tail-Optimal AllReduce for Distributed Deep Learning in the Cloud
Ertza Warraich, Omer Shabtai, Khalid Manaa, Shay Vargaftik, Yonatan Piasetzky, Matty Kadosh, Lalith Suresh, Muhammad Shahbaz
Google Research Scholar Award
PDF
Artifact
Computer Science and Engineering
Perfect is the enemy of good for distributed deep learning in the cloud
Leveraging deep learning’s resilience, approximating data lost by allowing some servers to time out speeds up model training while preserving performance.
Tech Xplore
Perfect is the enemy of good for distributed deep learning in the cloud
A new communication-collective system, OptiReduce, speeds up AI and machine learning training across multiple cloud servers by setting time boundaries rather than waiting for every server to catch up, ...
OSDI '24
Caravan: Practical Online Learning of In-Network ML Models with Labeling Agents
Qizheng Zhang, Ali Imran, Enkeleda Bardhi, Tushar Swamy, Nathan Zhang, Muhammad Shahbaz, Kunle Olukotun
SRC JUMP 2.0 Best Paper Award
PDF
Artifact
More …
Our Sponsors
NSF
National Science Foundation
SRC
Semiconductor Research Corporation
Intel
Intel Corporation
Google Research
Google
Facebook
Meta
VMware Research
by Broadcom
AMD
Advanced Micro Devices
Nvidia
Nvidia Corporation
ONF
Open Networking Foundation
Contact Us
University of Michigan
Computer Science and Engineering, EECS 
Location
Bob and Betty Beyster Building (4644), University of Michigan
Social Media
Connect with us on Twitter, LinkedIn, and GitHub