Seminars

Friday, February 6, 2026 
10:20 – 11:10 a.m. in ETB 1020

Dr. Ayobami S. Edun
Senior Assistant Vice President

Title: “AI in Highly Regulated Environments: From Data Integrity to Explainability, Benchmarking, and Model‑Risk Management”

Abstract:
AI systems in highly regulated environments must deliver reliability, transparency, and operational safety under demanding real‑world conditions. Particularly, AI models for fraud detection face extreme class imbalance, delayed ground‑truth labels, adversarial behavior, and continuous data drift—making rigorous validation both technically challenging and operationally essential. This seminar presents a reproducible, regulator‑aligned framework for assessing high‑risk AI/ML models, emphasizing data integrity checks, drift diagnostics, and performance stability testing. I will show how surrogate and offset models reconstruct global model behavior when vendor transparency is limited, how segment‑level diagnostics expose under‑performance, and how sensitivity and robustness analyses stress‑test stability under realistic perturbations. The talk concludes with practical strategies for automated validation pipelines, alert‑rate governance, and compensating controls that preserve model reliability while supporting frontline operations. The overarching goal is to illustrate how disciplined validation and model‑risk management enable trustworthy, compliant AI in tightly governed domains.

Biography:
Ayobami S. Edun, Ph.D., is a Senior Assistant Vice President focused on validating AI and machine‑learning systems used for fraud detection in highly regulated financial environments. His work spans statistical testing, model explainability, adversarial robustness, and regulatory compliance, with emphasis on developing reproducible and audit‑ready validation pipelines. He regularly leads the validation of high‑risk vendor and in‑house models, including neural‑network‑based real‑time fraud detectors and dynamic, frequently retrained systems whose opacity and operational constraints require surrogate modeling and automated testing frameworks.

Dr. Edun holds a Ph.D. in Computer Engineering and works extensively with Python, Spark, and secure MLOps tooling to enable robust, large‑scale evaluation of model behavior. His research interests include explainable AI, stress testing of ML systems, benchmarking under distribution shift, and the intersection of AI governance with statistical methodology.

Friday, January 30, 2026 
10:20 – 11:10 a.m. in ETB 1020

Dr. Lijun Qian
Texas A&M Regents Professor
Electrical and Computer Engineering
Prairie View A&M University

Title: “Task-Aware Design Optimizations for Analog In-Memory Inference Across Vision and Language”

Abstract:
Analog in-memory computing (AIMC) is a promising compute paradigm to improve speed and power efficiency of neural network inference beyond the limits of conventional von Neumann-based architectures. However, AIMC introduces fundamental challenges such as noisy computations and there lacks a comprehensive understanding of how analog inference generalizes across task domains, precision settings, and architectural scales, and how core design parameters such as cell bits, ADC resolution, and tile size jointly influence model reliability and efficiency. In this talk, we start with benchmarking inference performance of pretrained deep learning models on AIMC simulators. We conduct a comprehensive evaluation of analog inference across both vision and language tasks using three state-of-the-art simulators—CrossSim, AIHWKIT, and MemTorch. Then we systematically quantify how cell precision, ADC resolution, and crossbar tile size influence model accuracy, stability, and efficiency under realistic non-idealities. Results demonstrate that analog inference can achieve within 2–5% of digital baselines when parameters are tuned regarding to layer sensitivity and workload structure. Furthermore, we derive task-aware design guidelines recommendations for vision models and transformer-based NLP tasks based on our findings.

Biography:
Dr. Lijun Qian is a Texas A&M University System Regents Professor and an AT&T Endowed Professor in the Department of Electrical and Computer Engineering at Prairie View A&M University (PVAMU). He is also the Director of the Center of Excellence in Research and Education for Big Military Data Intelligence (CREDIT Center). He received his BE from Tsinghua University, MS from Technion-Israel Institute of Technology, and PhD from Rutgers University. Before joining PVAMU, he was with Bell-Labs Research in Murray Hill, New Jersey. He was a visiting professor at Aalto University, Finland. He led the CREDIT Center to win the first place in the AI tracks at Sea challenge organized by the US Navy, and the first place in the IEEE CyberC Big Data Competition organized by the IEEE Big Data Initiative. He received Best Paper Award in IEEE CAMAD 2025, AIxHEART 2025, IEEE RoboCom 2023, and IEEE Globecom 2017. His research interests are in the areas of big data processing, artificial intelligence, quantum information science and quantum machine learning, wireless communications and mobile networks, network security and intrusion detection, and computational and systems biology.

Blog: https://www.pvamu.edu/blog/lijunqian/

Friday, January 23, 2026
10:20 – 11:10 a.m. in ETB 1020

Dr. Pablo Rangel
Assistant Professor
Electrical Engineering
Texas A&M University–Corpus Christi

Title: “Exploration of Event-Based Vision Sensors for Autonomous Systems Applications”

Abstract:
At the Collaborative Robots and Agents Lab (CORAL) at TAMUC-CC we are currently exploring multiple applications of event-based camera sensors. We started by exploring the integration of dynamic vision sensors (DVS), and Real-Time Object Detection (RTOD) algorithms. DVS or neuromorphic cameras provide a solution by capturing motion as discrete “events” rather than full-frame images, offering high temporal resolution, low latency, and reduced data bandwidth. Produces visualization data with minimized complexity and throughput. Then, RTOD can be achieved with optimal resource usage accurately and at high speeds. An architecture or sensor fusion framework has been in development to integrate and make the most optimal use of combining diverse vision-based technologies. The framework was preliminary studied for the capabilities of event-based imagers combined with YOLO. Initial experimental work and observations were documented in the utilization of the V2E toolbox that generates realistic synthetic DVS events from intensity frames. The YOLOv12 algorithm is implemented in both the frame and its following raw synthetic DVS event images. At CORAL we currently obtained Prophesee PSK320MPCM2RPI5 GENX320MP-CM2 event-camera module with fixed M6 mount lens (FOV 104°)  and  GENX320MP-CCAM5 chip on board with interchangeable M12 mount lens (FOV 76°) for hands-on experiments to validate and further expand on experiments done with synthetic data. By observing these technologies, confounders and limitations is possible to plan into developing filtering and other documented solutions to fusion these computer vision tools. This research effort aims to develop advanced methods for real-time monitoring in fast changing environments such as autonomous systems-based reconnaissance, structural integrity inspections, law enforcement monitoring, collision avoidance, autonomy T&E and overall tracking, detection, and characterization of objects of interest.

Biography:
Pablo Rangel, Ph.D., is an Assistant Professor of Electrical Engineering at Texas A&M University–Corpus Christi, where he leads the Collaborative Robotic Agents Laboratory (CORAL). He earned his Ph.D. in Electrical and Computer Engineering from The University of Texas at El Paso in May 2017. Dr. Rangel’s teaching portfolio spans undergraduate and graduate courses in electrical and computer engineering, including Circuit Analysis, Electronic Systems Design, Communication Theory and Systems, Microprocessors and Microcontrollers, Electromagnetic Theory, and Dynamics and Control Systems. He has also developed and instructed laboratories in microcontroller design and circuit analysis, integrating experiential learning components to enhance student engagement and technical proficiency. His research focuses on test and evaluation of autonomous systems, unmanned aircraft systems, multi-agent robotics, sensor fusion, Internet of Things (IoT), artificial intelligence, and cyber-physical systems. His applied research extends to mechatronics, biomedical instrumentation, wireless communications, systems engineering, and cybersecurity. Dr. Rangel has collaborated on research sponsored by the U.S. Department of Defense and the National Science Foundation, serving as Co-Principal Investigator for NSF Award #2131263 and DoD Grant #W911NF-23-1-0186. He has also contributed to projects addressing spectrum management and radio telemetry analysis for the U.S. Army.

Homepage: https://www.tamucc.edu/files/php/views/faculty-details.php?profile=Pablo_Rangel

Friday, January 16, 2026
10:20 – 11:10 a.m. in ETB 1020

Dr. Yong Oh Lee
Assistant Professor
Industrial & Data Engineering
Hongik University, South Korea

Title: “Designing AI Pipelines for Multimodality in Data-Scarce Medical Systems”

Abstract:
Fully multimodal AI systems are often studied under the assumption of large, well-aligned datasets across modalities, an assumption that rarely holds in real medical environments. Many medical AI applications instead operate in regimes where data are limited, noisy, imbalanced, and only weakly aligned across sources. This talk examines data-scarce medical systems as concrete examples of this regime and focuses on the engineering considerations required to build robust AI pipelines under such constraints. Specifically, the talk begins with task-level model design and development, and then it examines a structural design direction that connects single-modality learning to multimodal expansion by leveraging segmentation outputs as intermediate representations and applying attention-based multi-view integration. These design choices show how system-level decisions shape model behavior and support gradual extension toward multimodality.

Biography:
Dr. Yong Oh Lee is an Assistant Professor in Industrial & Data Engineering at Hongik University, South Korea. He received his Ph.D. in Computer Engineering in 2012 from Texas A&M University. Before joining academia, he worked at Samsung Electronics in Korea and at the Korea Institute of Science and Technology Europe (KIST Europe) in Germany. His research focuses on artificial intelligence applications in healthcare, including multimodal medical imaging, clinical data integration, and large language models for medical information and drug safety. He has led projects on rare disease diagnosis, ophthalmology, orthopedics, oral cancer, and AI-enhanced educational systems. Dr. Lee actively collaborates with medical researchers to develop explainable and clinically reliable AI models that bridge engineering and medicine, aiming to advance data-driven healthcare innovation and support translational applications in clinical practice.

Homepage: https://sites.google.com/view/aiahongikuniversity/home

Aggie Graduate 2012! (Advisor: Dr. Reddy)

In conjunction with the ECE Leaders & Innovators:

Friday, Dec. 5, 2025
10:20 – 11:10 a.m. in ETB 1020

Tenured Associate Professor
Department of Materials Science and Engineering
Korea Advanced Institute of Science and Technology (KAIST)

Title: “Spatiotemporal Computing Utilizing Dual Thermal Dynamics of Mott Memristors”

Abstract:
The Mott memristor is a highly intriguing device that demonstrates unique electrical characteristics through the dynamic interaction of heat and current. The device exhibits dynamic thermal behavior, encompassing temporal accumulation via heat capacity and spatial transportation through heat diffusion. This spatiotemporal thermal activity enables coupling between memristor devices when arranged in arrays, which can be effectively utilized for computing. Additionally, the thermal dynamics of Mott memristors inherently involve stochasticity, resulting in probabilistic behavior. These properties, such as thermal coupling and stochasticity, provide a novel approach to tackling NP-hard problems, which are often challenging for conventional computers to solve. This presentation explores various computing devices that leverage the spatiotemporal thermal information of Mott memristors, including true random number generators (TRNGs), probabilistic computing systems, and thermal computing devices. The future potential and implications of these technologies will also be discussed.

Biography:
Professor Kyung Min Kim is a Tenured Associate Professor in the Department of Materials Science and Engineering at the Korea Advanced Institute of Science and Technology (KAIST) since 2017. He earned his B.S. degree in 2003 and his Ph.D. degree in 2008 from Seoul National University, Seoul, Korea. From 2011 to 2013, he worked at Samsung Electronics in Korea, and from 2014 to 2017, he worked at Hewlett Packard Labs of Hewlett Packard Enterprise in Palo Alto, California, USA. His research covers a wide range of areas related to next-generation semiconductor technology. This includes exploring new semiconductor materials and processing techniques, post-von Neumann computing technologies such as neuromorphic computing, reservoir computing, and probabilistic computing, as well as semiconductor packaging technology.

Friday, Nov. 21, 2025
10:20 – 11:10 a.m. in ETB 1020

PhD Student
Computer & Information Science at GRASP Lab
University of Pennsylvania

Title: “Communication and Perception as Enablers for Robust and Resilient Field Robotics”

Abstract:
Heterogeneous teams of aerial and ground robots are becoming increasingly mature, offering a synergistic approach in search and rescue, mapping, and surveillance applications. High-altitude UAVs can quickly map an environment as they fly in obstacle-free space. Maps can be transmitted to ground robots, which use them to validate objects of interest, thanks to a lower vantage point.

 In this talk, we will focus on two significant challenges for heterogeneous teams of robots in the field: communication and perception. First, we will explore opportunistic communications as a paradigm to facilitate large-scale deployment of aerial and ground robot teams. Second, we will discuss applications of neuromorphic sensors such as event cameras for field robotics, showing how these sensors excel in challenging lighting conditions encountered in field applications. Finally, we will address gaps for deploying heterogeneous teams of robots in the field at scale.

Biography:
Fernando Cladera is currently a Ph.D. student in Computer and Information Science at the GRASP Lab, University of Pennsylvania, Philadelphia, advised by Drs. Hsieh, Taylor, and Kumar. His research interests include event-based perception, field robotics, and micro and nano aerial vehicles. Fernando received an M.Sc. degree in Embedded Systems from the University of Rennes 1, Rennes, France, in 2013, and an M.S.E. degree in Robotics from the University of Pennsylvania in 2019. Before joining Penn, Fernando worked at the Samsung AI Center in NYC. Contact him at fclad@seas.upenn.edu.

Friday, Nov. 14, 2025
10:20 – 11:10 a.m. in ETB 1020

Gabe Dagani
Director of GPU Software
Samsung Austin Research Center

Title: “3D Rendering on the Edge: Developing Mobile GPUs for the Samsung Galaxy Phone”

Abstract:
A modern mobile phone puts the power of a game console in your pocket. This talk will review modern 3D rendering, and the challenges of making a flagship mobile phone like the Galaxy Flip7. No prior knowledge of 3D graphics is necessary.

Biography:
Gabe Dagani is a Director of GPU Software at Samsung Austin Research Center. He has over 20 years’ experience in 3D Graphics, video games, and embedded systems. His team is responsible for developing the software stack that goes into millions of phones every year to support all graphics rendering on Samsung Galaxy phones.

Friday, Nov. 7, 2025
10:20 – 11:10 a.m. in ETB 1020

Dr. Connor McCurley
Senior Machine Learning Specialist
Orbital Sidekick (OSK)

Title: “Translating Research into Real-World Impact: Hyperspectral Imaging, Orbital Sidekick, and Building a Career”

Abstract:
Hyperspectral imaging has emerged as a transformative technology in Earth observation, providing unprecedented spectral detail for applications spanning environmental monitoring, mineral exploration, and national security. This seminar will present an overview of hyperspectral imaging and its operational realization through Orbital Sidekick (OSK), a leader in on-orbit hyperspectral technology for global monitoring. We will examine common trends shaping the field, including advances in machine learning for spectral analysis, the integration of analytical workflows for actionable intelligence, and practical insights gained from real-world satellite missions. The seminar will conclude with an open discussion for graduate students navigating career decisions between academia and industry – emphasizing professional growth, translational research opportunities, and the application of advanced research skills in operational settings. Attendees will gain both a technical perspective on hyperspectral imaging and practical advice for leveraging their expertise in the evolving landscape of the professional world.

Biography:
Connor McCurley is a Senior Machine Learning Scientist at Orbital Sidekick (OSK).  He received the Ph.D. degree in Electrical and Computer Engineering from the University of Florida in Gainesville.  His research has primarily focused on developing machine learning methods to automatically understand and process data and imagery, with a specific focus on learning from imprecise, uncertain and ambiguous data annotations.  At OSK, Connor develops algorithmic methods for the improvement and exploration of hyperspectral imaging technologies. His work at OSK has included target detection, georeferencing, and generalized scene understanding.

Homepage Links:

• Orbital Sidekick: https://www.orbitalsidekick.com/
• Personal website: https://cmccurley.github.io/
• LinkedIn: https://www.linkedin.com/in/mccurleyconnor/

Friday, Oct. 31, 2025
10:20 – 11:10 a.m. in ETB 1020

Dr. Zigfried Hampel-Arias
Spectral and Geospatial Remote Sensing Team
Los Alamos National Lab

Title: “Machine Learning for Hyperspectral Imagery Applications: Trusting Our Models“

Abstract:
Hyperspectral imaging (HSI) comprises a rich source of remote sensing data and is crucial for remote detection of various targets, including gases, liquids, and solids. The advent of deep learning approaches has shown promising results for material detection and identification within long wave infrared (LWIR) HSI. We first introduce the utility of machine learning as a HSI analysis tool, explore the use of physics-guided neural networks as compared to black-box NN models for material classification, and discuss the role of explainability methodology in such analysis. We then discuss the application of 3D scene understanding methods using multiple observations from different angles to create spatially-relevant representations of HSI data, enhancing our ability to characterize targets when obscured by other objects in the scene. We will present results from our analysis pipeline, demonstrating our machine learning approach to the ill-posed inverse problem of material identification and the spectral scaling of neural radiance field (NeRF) models for 3D scene understanding, while also including uncertainty quantification methods to measure associated confidence in our spatial-spectral modeling.

Biography:
Zigfried Hampel-Arias is on the Spectral and Geospatial Remote Sensing team in the Intelligence and Space Research division at Los Alamos National Lab, with main research interests in applying deep learning to hyperspectral imaging, reinforcement learning to astro-dynamical systems and the national security implications of a robust innovation ecosystem. He received his BS in Chemical Physics from Rice University, conducted astrophysics research in Argentina, completed his PhD in physics at UW-Madison, finished a postdoc in Belgium and spent a few years in machine learning in the Bay Area. You can frequently find Zig somewhere in the mountains on bike or skis, occasionally with one of his daughters strapped tightly to him.

Friday, Oct. 24, 2025
10:20 – 11:10 a.m. in ETB 1020

Jarin Ritu
Ph.D. student, Department of Electrical & Computer Engineering
Texas A&M University

Title: “Texture-Informed Knowledge Distillation from Foundation Models for Efficient Deployment“

Abstract:
Large machine learning models achieve state-of-the-art performance but are often too computationally expensive for deployment in resource-constrained environments. To address this, researchers have turned to knowledge distillation (KD), a widely used method that transfers knowledge from large models into smaller, lightweight student models. However, conventional methods often fail to preserve fine-grained, domain-specific structures that are critical for robust generalization.
In this talk, we will introduce Statistical and Structural Audio Texture Knowledge Distillation (SSATKD), a framework that extends KD beyond prediction and feature-matching by embedding texture awareness into the transfer process. The framework incorporates modules to capture both structural patterns (e.g., periodicity, edges) and statistical properties (e.g., histograms, co-occurrence). By doing so, we enable compact student models to retain domain-relevant signal properties while remaining computationally efficient. We will present results showing that SSATKD improves performance over baseline KD methods in environmental sound tasks and demonstrate the framework’s adaptability to vision through segmentation of scanning electron microscopy images.

Biography:
Jarin Ritu is a third-year Ph.D. student in the Department of Electrical and Computer Engineering at Texas A&M University, working in the Advanced Vision and Learning Lab with Dr. Joshua Peeples. Her research lies at the intersection of machine learning and computer vision with applications in audio and image analysis, focusing on developing texture-informed knowledge distillation frameworks that transfer information from large foundation models to lightweight models for real-world deployment.
Her collaborative work with the Massachusetts Institute of Technology Lincoln Laboratory and Baylor University explores texture-informed AI methods to improve environmental sound analysis and quantify cellular morphology changes in microscopy images related to toxic exposures. She has presented her research at venues such as the IEEE Workshop on Applications of Signal Processing to Audio and Acoustics.
Beyond research, Jarin serves as the Vice President of Media for the Electrical and Computer Engineering Graduate Student Association at Texas A&M University and was selected for the National Science Foundation-funded Software-Tailored Architectures for Quantum Co-Design Quantum Ideas Summer School at Duke University. She also mentors undergraduate students in research projects related to her field.

Jarin Ritu’s homepage: avll.engr.tamu.edu/people/jarin-ritu/

Friday, Oct. 17, 2025
10:20 – 11:10 a.m. in ETB 1020

Chen Chen
Ph.D. student, Department of Electrical & Computer Engineering
Texas A&M University

Title: “Hardware Fuzzing to Secure Modern Hardware“

Abstract:
Hardware is at the heart of computing systems. However, recent years have seen increased attacks exploiting hardware vulnerabilities and exploits, which even traditional software-based protections cannot prevent. Hardware fuzzing has shown promise in detecting vulnerabilities in large-scale designs, such as modern processors. In this talk, I will first introduce hardware fuzzing as a method for finding vulnerabilities and its limitations. Then, I will detail how fuzzing techniques can be integrated with existing methods, such as formal verification and information flow tracking, to tackle these limitations. Finally, I will discuss improving the efficiency of hardware fuzzing by leveraging knowledge from prior processor generations and future works. 

Biography:
Chen Chen is a Ph.D. student in the Department of Electrical and Computer Engineering at Texas A&M University, supervised by Dr. JV. Rajendran. He obtained a B.S. degree from Purdue University. His research interest is hardware security. He has published papers at top-tier conferences like USENIX Security, DAC, ICCAD, and DATE. He has won the Distinguished Paper Award and the Distinguished Artifact Evaluation Reviewer Award at USENIX Sec’24. He also co-organizes HackTheSilicon, the world’s largest hardware security capture-the-flag competitions, co-located with DAC, USENIX Security, CHES, and DATE.

Chen Chen’s homepage: www.chenc.contact/

Friday, Oct. 10, 2025
10:20 – 11:10 a.m. in ETB 1020

Dr. Zhiwen Fan
Electrical & Computer Engineering Assistant Professor
Texas A&M University

Title: “Learning the World in 3D: Real-Time Algorithms and Scalable Data“

Abstract:
3D foundation models are a key step toward spatial and physical intelligence, yet their impact is often constrained by heavy computation and scarce, well-curated 3D data. In this talk, I will present a family of 3D algorithms that make reconstruction, rendering, and generation practical on real workloads, together with a scalable 3D data engine spanning human avatars, Martian-surface simulation, and cryo-EM microscopy.

First, I will cover: (1) city-scale 3D reconstruction and a real-time 3D perception Transformer that operates from uncalibrated images; (2) efficient neural rendering techniques that achieve 60% speedup over 3D Gaussian Splatting while preserving fidelity; and (3) methods by which explicit 3D understanding augments 2D vision–language models to improve spatial awareness prior to reasoning.

Next, I will introduce systematic remedies for 3D data scarcity: (1) an automated pipeline for multimodal human-avatar and environment annotation; (2) a reliable system for simulating Martian surfaces under limited texture availability; and (3) a learned inverse Fourier-slicing map that enables cryo-EM reconstructions up to 10× faster without sacrificing accuracy.

Taken together, these advances move us toward AI systems that interact with the physical world with genuine spatial understanding and real-time performance.

Biography:
Zhiwen (“Aaron”) Fan is an Assistant Professor in the Department of Electrical and Computer Engineering at Texas A&M University. He received his Ph.D. from The University of Texas at Austin. He was awarded the 2022 Qualcomm Innovation Fellowship and a Best Paper award at the CVPR AI4CC Workshop. He has served as an Area Chair for multiple AI/ML conferences and has completed research internships at Meta, NVIDIA, and Google.

Zhiwen Fan’s homepage: zhiwenfan.github.io

Friday, Sept. 26, 2025
10:20 – 11:10 a.m. in ETB 1020

Dr. Saurabh Bagchi
School of Electrical & Computer Engineering
Dept. of Computer Science
Purdue University

Title: “Dependability and Data Analytics: A Grand Challenge of Our Times”

Abstract:
We live in a data-driven world as we have been realizing for some time. Everything is generating data, in volumes and at high rates, from the sensors embedded in our physical spaces to the large number of machines in data centers being monitored for a variety of metrics. The question that I pose in this talk is:

     Can all this data be used for improving the dependability of our computing systems?

Dependability is the property that a computing system continues to provide its functionality despite the introduction of faults, either accidental faults (design defects, environmental effects, etc.) or maliciously introduced faults (security attacks, external or internal). We have been addressing the dependability challenge through large-scale data analytics applied end-to-end from the small (networked embedded systems, mobile and wearable devices) [e.g., MobiSys-25, Security & Privacy-24, CVPR-24, 23, 22, TODAES-23, Eurosys-22] to the large (edge and cloud systems, distributed machine learning clusters) [e.g., NeurIPS-25, 22, UAI-25, CCS-24, Security & Privacy-24, ICLR-24, Middleware-23, OSDI-22, Sigmetrics-22]. In this talk, I will first give a high-level view of how data analytics has been brought to bear on dependability challenges, and key insights and achievements from our technical community. Then I will provide details of three projects where we design systems from the small to the large, to meet dependability and performance requirements, using data-driven decisions.

For the detailed part, I will discuss the following projects: (I) Small: Streaming LiDAR analytics on autonomous vehicles; (II) Small-to-Medium: Sensor-edge-cloud continuum; (III) Small-to-Medium: Secure decentralized learning. I will provide enough background and intuition in each, so that the talk is accessible to a broad CS audience.

Biography:
Saurabh Bagchi is a professor in the School of Electrical and Computer Engineering and the Department of Computer Science at Purdue University in West Lafayette, Indiana. His research interest is in dependable computing and distributed systems. He is the Director of the NSF CISE Center Chorus (2024-ongoing) and of the Army’s Artificial Intelligence Innovation Institute (A2I2) (2020-ongoing). He was selected to the International Federation for Information Processing (IFIP) (2020) and is a Fellow of the Institute of Engineering and Technology (IET) (2022). Saurabh serves on the Board of Governors of the IEEE Computer Society and is the Founder of a cloud computing startup, KeyByte.

Saurabh is proudest of the 25 PhD students and 50 Masters thesis students who have graduated from his research group and who are in various stages of building wonderful careers in industry or academia. In his group, he and his students have way too much fun building and breaking real systems. Along the way, this has led to 13 best paper awards or runners-up awards at IEEE/ACM conferences and a Test of Time Award. Saurabh received his MS and PhD degrees from the University of Illinois at Urbana-Champaign and his BS degree from the Indian Institute of Technology Kharagpur, all in Computer Science.

Dr. Saurabh Bagchi’s homepage is https://www.saurabhbagchi.us.

Friday, Oct. 3 2025
10:20 – 11:00 a.m. in ETB 1020

Marcus Botacin
Computer Science Assistant Professor
Texas A&M University

Title: “Hardware is the New Software: The Next-Gen Anti-Viruses and How Your Hardware Will Self-Secure Your System!”

Abstract:
New threats require new defenses. In the same way that LLMs are reshaping malware creation, hardware support for security should be reshaping defenses. In this talk, Dr. Botacin presents the proposal for the migration of AV engines to hardware to implement continuous monitoring systems (Zero-Trust) with low performance impact. The presentation incrementally complexifies hardware design, from external FPGA AVs to the redesign of the memory controller. The talk discusses ideas like adding new Memory Management Unit (MMU) flags to create a new type of page fault – the security fault – to create security-aware OSes that invoke AntiViruses (AVs) on demand to scan processes in more suitable states. Dr. Botacin’s research talk will highlight how AVs can be significantly enhanced by relying on additional hardware support and will call for action to advance their development.

Biography:
Dr. Botacin is a Computer Science Assistant Professor at Texas A&M University (TAMU, USA) since 2022. Ph.D. in Computer Science (UFPR, Brazil), master’s in computer science and computer engineering (UNICAMP, Brazil). Malware Analyst since 2012. Specialist in AV engines and Sandbox Development. Dr. Botacin published research papers at major academic conferences and journals. Dr. Botacin also presented his work at major industry and hacking conferences, such as HackInTheBox and Hou.Sec.Con.

Marcus Botacin’s homepage: https://marcusbotacin.github.io/

Friday, Sept. 19 2025
10:20 – 11:10 a.m. in ETB 1020

Varun Murali
Assistant Professor, Electrical & Computer Engineering
Texas A&M University

Title: “From Robots that React to Robots that Adapt”

Abstract:
Achieving truly autonomous robotic systems that can adapt to unknown environments and collaborate using natural language requires a fundamental shift beyond the traditional “sense-think-act” paradigm. In this talk, I will advocate for a holistic framework centered on active perception, where perception, reasoning, and action are tightly integrated into a unified loop. This approach is key to developing resilient, adaptive robots capable of operating in novel and unstructured settings. We will explore the interplay between perceptual geometry, semantics, and action within this paradigm. The first part of the talk explores perception-aware planning and the necessity for coupling the problem for perceptual degradation. The latter half focuses on how foundation models for embodied AI can enable scalability and generalization. A crucial aspect of this discussion is the challenge of translating natural language into actionable information — determining what to map, which areas to explore, and how to represent knowledge in a way that remains scalable while retaining task relevance. Additionally, I will discuss hierarchical architectures that support effective decision-making across diverse contexts and how to build guardrails on inferred actions from foundation models to ensure responsible and safe autonomy. By addressing these challenges, we move toward adaptive, resilient, and responsible robotic systems that can function collaboratively with humans and other robots at scale. This vision lays the groundwork

Biography:
His research develops computationally efficient algorithms for hierarchical semantic mapping, planning, and tasking, enabling robots to operate in dynamic and unstructured environments. Prior to joining Texas A&M, he was a Postdoctoral Researcher at the GRASP Laboratory, University of Pennsylvania, with Prof. Vijay Kumar, and earned his Ph.D. in Aeronautics and Astronautics from MIT under Prof. Sertac Karaman. His work emphasizes general-purpose autonomy frameworks that tightly integrate perception and action to improve decision-making and adaptability. He has contributed methods for semantic reasoning, robust localization and mapping, and perception-aware planning in challenging settings. Currently, he is exploring how foundation models and large language models can advance task representations, safe autonomy, and size, weight, and power-constrained robotic systems.

Varun Murali’s Homepage: https://varunmurali1.github.io

Friday, Sept. 12 2025
10:20 – 11:10 a.m. in ETB 1020

Stephen McCullough
Alumnus from Texas A&M Engineering
Dept. of Electrical Engineering

Title: “Solution Engineering, Explained”

Abstract:
This seminar will illuminate the world of solutions engineering and why it may be the right path for you. You’ll learn about careers in solutions engineering, how the role differs from conventional engineering positions, and the skills that set successful solutions engineers apart.

Audience:
Engineering students of all years and disciplines welcomed!

Biography:
Stephen McCullough graduated from Texas A&M with a B.S. in Electrical Engineering, and he has since built his career as a solutions engineer. He has worked at Dell Technologies, Amazon, and now NVIDIA, where he designs and implements large-scale AI infrastructure deployments for cloud providers. Stephen has gained valuable insights from these experiences, and he is eager to share his knowledge with Aggie engineering students.

Stephan McCullough’s Homepage: https://www.linkedin.com/in/stephen-mcc

Friday, Sept. 5 2025
10:20 – 11:10 a.m. in ETB 1020

Fabia Farlin Athena
Energy Postdoctoral Fellow in Electrical Engineering
Stanford University

Title: “Emerging Materials and Devices for Energy-Efficient Nanoelectronics”

Abstract:
The rapid rise of data-intensive workloads, driven by artificial intelligence, is exposing the fundamental energy and latency limits of conventional computing hardware. The memory wall bottleneck, in particular, demands a shift toward new hardware architectures designed to minimize costly data movement. In this seminar, I will present advances at the intersection of materials science and device engineering that address this challenge. First, I will discuss adaptive oxide-based resistive memories for in-memory computing, detailing how novel MAX-phase electrodes achieve record-low off-state currents and how an in-situ recovery method enables robust transfer learning on analog AI accelerators. Next, I will introduce complementary back-end-of-line technologies that advance scalability for high-density, on-chip memory, including oxide semiconductor gain-cells and an interface-dipole engineering strategy for precise threshold voltage tuning. Together, these results highlight a practical and scalable path toward vertically integrated, energy-efficient electronics for future data-centric applications.

Biography
Dr. Fabia Farlin Athena is an Energy Postdoctoral Fellow in Electrical Engineering at Stanford University, working with Prof. H.-S. Philip Wong and Prof. Alberto Salleo to develop high-bandwidth oxide-semiconductor gain-cell memories and monolithic 3-D memory/logic stacks for ultra-low-power AI systems. She earned her Ph.D. and M.S. in Electrical & Computer Engineering from Georgia Institute of Technology, advised by Prof. Eric M. Vogel, where her research focused on adaptive oxide memristors for brain-inspired AI, receiving the Sigma Xi Best Ph.D. Thesis Award. She has also held research scientist intern positions at IBM TJ Watson. Fabia’s interdisciplinary contributions have been recognized with the IBM PhD Fellowship, MRS Graduate Student Award, Cadence Technology Scholarship, VLSI Symposium Highlight Paper, EECS Rising Stars, Stanford Energy Postdoctoral Fellowship, and Forbes 30 Under 30 North America.

Dr. Athena’s Homepage: https://ffathena.github.io/

Friday, April 11, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020

Dr. Kaushik Chowdhury 
Chandra Family Endowed Distinguished Professor
Dept. of Electrical Engineering & Computer Engineering
The University of Texas, Austin

Title: “Programming the Network and the Environment for Sensing, Communication and Computation”

Abstract:
This presentation delves into how extreme reconfigurability will shape future wireless networks, made possible by exciting new developments in the areas of open radio access networks (O-RAN) and reconfigurable intelligent surfaces (RIS). By embracing programmability at all levels, from centralized cellular architectures to distributed network components, such wireless networks will not only establish resilient communication links but also perform sensing tasks to better perceive and, to an extent, even engineer the surrounding environment. In this two-part talk, we first describe how “programmable networks” in the form of O-RAN-compliant base stations can evolve their roles from sensing weak radar signals to performing automated network traffic analysis followed by resource allocation. In the second part of the talk, we shift towards “programmable environments” where we experimentally demonstrate how custom-designed RIS can be used for channel hardening in single antenna receivers by intentionally creating multipath diversity. We also explore how such programmability opens a new domain of wireless computing, where transmitted signals can be used to perform mathematical operations by leveraging the physics of over-the-air propagation. A common theme in programming both the “network” and “environment” is rigorous experimental validation on laboratory and field-deployed systems.

Biography
Kaushik Chowdhury is a Chandra Family Endowed Distinguished Professor in the Department of Electrical & Computer Engineering at The University of Texas at Austin. His primary interest lies in the areas of applied machine learning for wireless, including signals intelligence, spectrum monitoring and sharing in commercial and dual-use environments, dataset generation, and network optimization by harnessing big data. Prof. Chowdhury has worked on several large-scale wireless community infrastructure projects that include the Colosseum RF/network emulator as well as the Platforms for Advanced Wireless Research project office. Prof. Chowdhury was a finalist for the 2023 US Blavatnik National Awards for Young Scientists. He was also the winner of the U.S. Presidential Early Career Award for Scientists and Engineers (PECASE) in 2017, the Defense Advanced Research Projects Agency Young Faculty Award in 2017, the Office of Naval Research Director of Research Early Career Award in 2016, and the National Science Foundation (NSF) CAREER award in 2015. He is an IEEE Fellow and ACM Distinguished Member.

Friday, March 25, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020

Dr. Vijay K. Shah, PhD
Assistant Professor
Dept. of Electrical Engineering and Computer Engineering
North Carolina State University

Title: “Advancing the Future of Telecoms with Open RAN”

Abstract Open Radio Access Network (Open RAN) is a transformative technology that is reshaping the global telecom landscape. At its core, Open RAN is built on two fundamental principles — openness and intelligence — which are enabled by advancements in virtualization, cloudification, softwarization, programmability, and open interfaces. Together, these innovations enable the integration of AI-powered, automated, and closed-loop RAN control functions, significantly enhancing the deployment and management of cellular networks. This presentation will explore emerging trends in 5G/6G telecom networks, with a particular focus on the evolution of Open RAN. We will examine recent research works from the NextG Wireless Lab at NC State, including notable RAN control applications, namely, Interference Classification xApp, IMPACT xApp, Radar Detection xApp, ORANSight — the foundational LLMs for O-RAN, among others. The talk will conclude with a forward-looking discussion on Open RAN’s critical role in shaping the future of telecom systems, establishing it as a cornerstone for the development of next-generation networks.

Biography Dr. Vijay K. Shah is an Assistant Professor in the Electrical and Computer Engineering (ECE) Department at NC State University, where he leads the NextG Wireless Lab, a cutting-edge research group focused on pioneering innovations in next-generation wireless communication and networking. Dr. Shah is also the co-founder and CEO of WiSights, a spinoff company advancing large language model (LLM) solutions for next-generation telecom networks. In addition, he serves as the outreach director for AERPAW (Aerial Experimentation and Research Platform for Advanced Wireless), the nation’s first research platform dedicated to advancing wireless technologies and autonomous drone systems. Dr. Shah’s research spans several critical domains, including 5G/6G networks, Open RAN, spectrum sharing and AI/GenAI for wireless networks. His work is supported by various federal and state agencies, including the NSF, NIST, and NTIA. Dr. Shah is a co-author of “Fundamentals of O-RAN”, the first comprehensive book on Open RAN technology.

 Shah’s personal webpage: https://ece.ncsu.edu/people/vkshah2/

Friday, April 4, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020

Zhongyuan Zhao
Research Assistant Professor
Dept of Electrical Engineering and Computer Engineering
Rice University

Title: “Distributed AI in Networked Systems: A Graph-Based Neuro-Symbolic Perspective”

Abstract Artificial intelligence has achieved remarkable success in processing regular, unstructured data, such as language and images, where information is supported on structured grids (tokens in 1D sequences, pixels in 2D grids). However, networked systems—such as communication networks, transportation and logistics systems, and fog/edge/cloud computing—pose fundamentally different challenges for standalone AI. These systems involve parallel and rule-based operations, irregular and dynamic structures, and real-time, high-stakes decision-making, rendering standalone AI systems insufficient. This talk explores graph-based neuro-symbolic approaches that incorporate domain knowledge to enable AI adhere to rules and protocols, learn and adapt with smaller models and less training data, while remaining scalable and interpretable. I will first introduce graph neural networks (GNNs) as a universal workhorse for learning from graph-structured data. Next, we will dive into graph-based deterministic policy gradient (GDPG), a general framework that enhances rather than replaces domain-specific heuristics with GNNs. Using link scheduling in self-organizing wireless networks as an example, we will explore how this approach can tackle key challenges in networked systems—real-time constraints, combinatorial decision spaces, lack of labeled data, and the need for scalability and interpretability. This talk will conclude with an overview of recent advances, potential applications, and broader implications of this framework, including its relevance to biological networks (e.g., protein interaction networks), adaptive traffic signal control, and knowledge graphs.

Biography Dr. Zhongyuan Zhao is a Research Assistant Professor in the Department of Electrical and Computer Engineering at Rice University, working at the intersection of machine learning, network science, wireless communications, and operations research. His research focuses on graph-based neuro-symbolic approaches that integrate the structural, organic, and engineered dimensions of complex, networked systems. By bridging graph-based machine learning and domain-specific analytical models, he has advanced distributed AI architectures, combinatorial optimization, stochastic network optimization, and digital signal processing, leading to scalable and interpretable intelligent solutions for edge computing, routing, scheduling, and baseband signal processing in wireless networks. His work has been published in leading venues in machine learning (ICLR), wireless communications (IEEE TWC, JSAC, TMLCN, etc.), signal processing (IEEE ICASSP, CAMSAP), operations research (M&SOM), and bioinformatics (Brief. Bioinform.). Dr. Zhao collaborates with ARL, USC-ECE, and UNL-Biochem. He has served as Session Chair at ICASSP (2022, 2023) and as a TPC member for IEEE VTC (2020) and Asilomar (2025). He is also a recipient of the Future Faculty Fellowship at Rice University. Dr. Zhao earned his Ph.D. in Computer Engineering from the University of Nebraska-Lincoln, and holds an M.S. in Signal Processing and B.S. in Information Countermeasures Technology from the University of Electronic Science and Technology of China.

 Dr. Zhongyuan Zhao’s personal webpage: https://zhongyuanzhao.com/

Friday, March 21, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Xiongye Xiao
PhD Candidate , Dept. of Electrical & Computer Engineering
University of Southern California

Title: “AI for Science: From Microscopic Structures and Dynamics to Macroscopic Functions”

Abstract
The evolution of scientific discovery has transitioned from observational studies and mathematical modeling to data-driven AI methodologies. With the rapid advancement of deep learning and neural network architectures, AI has become a powerful tool for accelerating scientific discovery. However, as large-scale models exhibit emergent behaviors, understanding the structural and dynamic principles that govern these models is crucial.

In this talk, I will discuss Neuron-based Multifractal Analysis (NeuroMFA), a novel framework that bridges AI and scientific discovery by characterizing the self-organization of large models. Inspired by neuroscience, NeuroMFA provides a structural perspective on emergent intelligence in AI, offering insights into how deep neural networks develop complex functionalities. This approach enables a deeper understanding of AI’s role in science and facilitates advancements in areas such as material discovery, physics-informed machine learning, and computational neuroscience.

Biography
Xiongye Xiao is a researcher specializing in AI for Science, neural operators, and network science. He is a Ph.D. candidate in Electrical and Computer Engineering at the University of Southern California. His work focuses on developing advanced mathematical and machine learning frameworks to bridge microscopic structures and dynamics to macroscopic functions, with applications in neuroscience, material science, and AI-driven scientific discovery. He has published in leading venues such as NeurIPS, ICLR, and Science Robotics, and his contributions include pioneering Neuron-based Multifractal Analysis (NeuroMFA) for analyzing emergent intelligence in large models, as well as the Multiwavelet Neural Operator, which enhances PDE learning through multi-resolution representations.

Please join us on Friday, 3/21/25 at 10:20 a.m. in ETB 1020!

Xiongye Xiao’s Webpage: xiongyexiao.com
Google Scholar Page: https://scholar.google.com/citations?user=AvIxA64AAAAJ&hl=en&oi=ao

Host: Dr. Jiang Hu

Friday, March 7, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Cunxi Yu
Assistant Professor, Dept. of Electrical & Computer Engineering
University of Maryland, College Park

Title: “The Rise and Fall of Machine Learning for EDA and Beyond – Studies in Synthesis and Verification”

Abstract
In recent years, Machine Learning (ML) has gained considerable momentum in electronic design automation (EDA). Specifically, the successes of ML-driven EDA methods and infrastructure have demonstrated its unique capability in capturing the multitude of factors affecting estimation accuracy, effectively exploring large algorithmic and design spaces in synthesis, and accelerating classical combinatorial optimization problems. In particular, synthesis and verification, critical stages in EDA, have significantly benefited from ML in the last five years. However, during the development of ML-driven synthesis and verification approaches, several points of convergence have been observed, including practicality, system engineering challenges, data availability, and determinism. In this talk, I will present the journey of exploring ML in synthesis and verification, focusing on discussing the evolutionary developments from static ML-based synthesis approaches to algorithmic learning and general combinatorial optimizations using advanced domain-specific ML techniques. The talk will primarily focus on our recent work in high-level synthesis, logic synthesis, and Boolean reasoning.

Biography
Dr. Cunxi Yu is an Assistant Professor at the University of Maryland, College Park. His research interests center around novel algorithms, systems, and hardware designs for computing and security. Before joining the University of Maryland, Cunxi was an Assistant Professor at the University of Utah and held a PostDoc position at Cornell University. His work has received the Best Paper Award at DAC (2023), the NSF CAREER Award (2021), American Physical Society DLS poster award (2022), and multiple best paper nominations. Cunxi earned his Ph.D. from UMass Amherst in 2017.

Please join us on Friday, 3/7/25 at 10:20 a.m. in ETB 1020!

Dr. Yu’s personal webpage and Google Scholar pages are linked from here: https://ece.umd.edu/clark/faculty/1834/Cunxi-Yu

Host: Dr. Jiang Hu

Friday, Feb. 21, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Yu Wang
PhD Candidate, Dept. of Computer Engineering
University of California, Santa Barbara

Title: “Enable Efficient Bayesian Optimization with Semi-Supervised Learning and In-Context Learning of Large Language Models”

Abstract
Bayesian Optimization (BO) is a powerful framework for optimizing expensive black-box functions, but its efficiency is often hindered by two key challenges: (1) difficulties in encoding domain knowledge and (2) data scarcity that limits surrogate model generalization. In this talk, I will present the recent efforts from our group to address these challenges.

First, we introduce ADO-LLM, which leverages the domain priors of Large Language Models (LLMs) to assist BO in tackling the challenging analog circuit sizing problem. By proposing feasible and high-potential design parameters, LLMs enable BO to efficiently explore design spaces while balancing complex tradeoffs between design specifications.

Second, we discuss TSBO, a method that enhances BO’s data efficiency in high-dimensional settings by strategically incorporating unlabeled samples and generating reliable pseudo-labels via a teacher-student model with feedback. TSBO demonstrates strong performance across multiple high-dimensional BO tasks, achieving up to a 364.2x reduction in labeled data usage.

Finally, I will conclude with a discussion on the broader implications and future directions for data-efficient BO.

Biography
Yu Wang is a Ph.D. candidate in Computer Engineering at the University of California, Santa Barbara, advised by Professor Peng Li. He received his M.S. from Texas A&M University in 2019 and his B.S. from Fudan University.

Yu’s research focuses on the intersection of advanced machine learning and hardware system design, with an emphasis on Bayesian Optimization, Semi-Supervised Learning, and hallucination mitigation in Large Language Models (LLMs). He explores their applications in electronic design automation (EDA) to enhance design efficiency and automation.

His works have been recognized at leading machine learning and circuit design venues, including AAAI, ICML, TMLR, and ICCAD. He also received the Best Paper Award at ASAP 2020.

Please join us on Friday, 02/21/25 at 10:20 a.m. in ETB 1020 to learn more on the research presented by CESG’s former M.S. ECE graduate Yu Wang!

Host: Alex Sprintson; Faculty may request a Zoom Link for this presentation.

Wang’s Google Scholar: https://scholar.google.com/citations?user=lUd8s0QAAAAJ&hl=en

Friday, Feb. 28, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Yiorgos Makris
Professor, Dept. of Electrical and Computer Engineering
University of Texas, Austin

Title: “Provenance Attestation: From Silicon Chips to Biological Cells and Beyond”

Abstract
Complex processes, whether natural or artificial, often exhibit inherent variability and result in slightly different products even when identical steps, equipment, materials and conditions are employed. Such variability typically consists of a random component, which is attributed to the endogenous stochasticity of the process itself, and a systematic component, which is attributed to the exogenous aspects of the production. In this presentation, we will discuss how this variability can be harnessed for the purpose of attesting both the process and each copy of the product, thereby facilitating trust, traceability and intellectual property protection. First, in the context of semiconductor manufacturing, using both physical and electrical measurements (a.k.a., metrology and wafer acceptance tests, respectively) from wafers manufactured using multiple copies of a mask-set in a 12nm GlobalFoundries technology, we will demonstrate the use of contemporary statistical and machine learning-based methods for determining whether a wafer was produced by a ratified mask-set. Leveraging the insight gained through this analysis, we will also discuss the design of custom sensors for obtaining the relevant information from each die on a wafer to collectively attest the mask-set used to produce this wafer, without relying on potentially untrusted foundry-provided data. Then, in the context of synthetic biology, using amplicon sequencing data from multiple cell lines (i.e., HEK293, HCT116 and HeLa), we will demonstrate that the stochasticity of the non-homologous end-joining (NHEJ) DNA repair process can be leveraged as a mechanism for introducing a unique identifier (i.e., a Genetic Physical Unclonable Function (PUF)) in every legitimately produced copy of a cell line. Akin to their counterparts in the semiconductor industry, Genetic PUFs can be used for attesting the provenance and protecting the intellectual property of valuable, genetically-engineered cell lines.

Biography
Yiorgos Makris received the Diploma of Computer Engineering from the University of Patras, Greece, in 1995 and the M.S. and Ph.D. degrees in Computer Engineering from the University of California, San Diego, in 1998 and 2001, respectively. After spending a decade on the faculty of Yale University, he joined UT Dallas where he is now a Professor of Electrical and Computer Engineering, the Co-Founder and Site-PI of the NSF Industry University Cooperative Research Center on Hardware and Embedded System Security and Trust (NSF CHEST I/UCRC), as well as the Leader of the Safety, Security and Healthcare Thrust of the Texas Analog Center of Excellence (TxACE) and the Director of the Trusted and RELiable Architectures (TRELA) Research Laboratory. His research focuses on applications of machine learning and statistical analysis in the development of trusted and reliable integrated circuits and systems. He has served as an Associate Editor of the IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, the IEEE Transactions on Information Forensics and Security and the IEEE Design & Test of Computers Periodical, and as a guest editor for the IEEE Transactions on Computers and the IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems. He has also served as the 2016-2017 General Chair and the 2013-2014 Program Chair of the IEEE VLSI Test Symposium. He is a recipient of the 2006 Sheffield Distinguished Teaching Award, Best Paper Awards from the 2013 IEEE/ACM Design Automation and Test in Europe (DATE’13) conference and the 2015 IEEE VLSI Test Symposium (VTS’15), as well as Best Hardware Demonstration Awards from the 2016 and the 2018 IEEE Hardware-Oriented Security and Trust Symposia (HOST’16 and HOST’18) and a recipient of the 2020 Faculty Research Award from the Erik Jonsson School of Engineering and Computer Science at UT Dallas.. Yiorgos is a Class of 2025 IEEE Fellow, for contributions to machine-learning-based design of trusted and reliable integrated circuits.

For more on Dr. Yiorgos Makris visit his website at https://personal.utdallas.edu/~gxm112130/.

Please join us on Friday, 2/28/25 at 10:20 a.m. in ETB 1020 to learn more and meet Dr. Yiorgos Makris!

Friday, Feb. 14, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Ulisses Braga-Neto 
Professor, Dept. of Electrical and Computer Engineering
Texas A&M University

Title: “DeepOSets: Non-Autoregressive In-Context Learning of Supervised Learning Operators” 

Abstract
We introduce DeepSets Operator Networks (DeepOSets), an efficient, non-autoregressive neural network architecture for in-context learning of permutation-invariant operators. DeepOSets combines the operator learning capabilities of Deep Operator Networks (DeepONets) with the set learning capabilities of DeepSets. Here, we present the application of DeepOSets to the problem of learning supervised learning algorithms, which are continuous permutation-invariant operators. We show that DeepOSets are universal approximators for this class of operators. In an empirical comparison with a popular autoregressive (transformer-based) model for in-context learning of linear regression, DeepOSets reduced the number of model weights by several orders of magnitude and required a fraction of training and inference time, in addition to significantly outperforming the transformer model in noisy settings. We also demonstrate the multiple operator learning capabilities of DeepOSets with a polynomial regression experiment where the order of the polynomial is learned in-context from the prompt.

Biography
Ulisses Braga-Neto received his Ph.D. in Electrical and Computer Engineering from The Johns Hopkins University in 2002. He is currently a Professor in the Electrical and Computer Engineering Department at Texas A&M University. His research focuses on Machine Learning and Statistical Signal Processing. Dr. Braga-Neto is the founding Director of the Scientific Machine Learning Lab at the Texas A&M Institute of Data Science (TAMIDS). He has published two textbooks and more than 160 peer-reviewed journal articles and conference papers. Dr. Braga-Neto received the NSF CAREER Award in 2009.

Please join us on Friday, 02/14/25 at 10:20 a.m. in ETB 1020 to learn more and meet our own Dr. Ulisses Braga-Neto!

For more on Dr. Ulisses Braga-Neto visit his website at https://cesg.tamu.edu/faculty/ulissess-braga-neto/.

Friday, Feb. 7, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Venkat Arun
Assistant Professor, Dept. of Computer Science
University of Texas, Austin

Title: “Synthesizing Provably Performant Controllers for Networked Systems” 

Abstract
Computer systems rely on various controllers to allocate CPU, memory, network and storage resources. The quality of their decision making significantly impacts system performance, sometimes degrading it by over 10x. Poor controller decisions also lead to performance variability that makes it difficult to build real-time applications and highly distributed computations. Controller design is hard due to incomplete system observability and the lack of an accurate system model. This makes it difficult to formally define, let alone guarantee, desired performance properties. As a result, these controllers are designed largely through empirical trial-and-error.

In this talk, I will discuss a set of principles and tools that we have developed to not only verify, but also synthesize controllers with provable performance guarantees. I will focus on the controllers involved in low-latency video streaming and show how program synthesis techniques can uncover novel solutions overlooked by human designers for decades.

Biography
Venkat Arun is an assistant professor at UT Austin. His research seeks to enable networked systems to support the next generation of applications. His past work has spanned internet congestion control, video streaming, privacy-preserving computation, wireless networks, and mobile systems. A common theme across these is the use of theoretical ideas to gain insights into the real-world that would have been difficult to discover otherwise. He has received multiple dissertation and best paper awards and the Marconi society young scholar award. He received his PhD at MIT.

Friday, Jan. 31, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020

Dr. Yale Patt
Professor, Electrical and Computer Engineering 
University of Texas, Austin

Title: “Computer Architecture Research: Moving the Needle” 

Abstract
After almost 60 years in the trenches performing microarchitecture research, I have some thoughts about what it takes to move the needle.  Most importantly, one must understand the hardware details that the microarchitecture is built on, and the details of the resulting microarchitecture.  Re: bottom-up or top-down, at least avoid being a top-up.  If you have a burning desire to add a feature, choose the microarchitecture, not the ISA. Understand heterogeneity and why it is almost certainly the answer.  Celebrate MIT’s proclamation that there is plenty of room at the top and my suggestion that there is still plenty of room at the bottom.  Recognize that there may be more than one obstacle in place so your improvement may not pay off until the last obstacle has been conquered.  Finally, be generous in your praise for the work of others.  You are not the only smart person in the room.  In this talk, I hope to expand on many of the points mentioned above.

Biography
Dr. Yale Patt is a teacher at UT and the Virginia Cockrell Centennial Chair in the Cockrell School of Engineering.  He earned obligatory degrees at reputable universities and has received more than his share of awards for his research and teaching.

More details on Dr. Patt’s webpage HERE.

Please join us on Friday, 01/3125 at 10:20 a.m. in ETB 1020 to glean some of his wisdom and to meet Dr. Yale Patt!

Friday, Jan. 24, 2025
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Toros Arikan
Postdoctoral Researcher, Electrical and Computer Engineering Dept.
Rice University

Title: “Deep Learning for Smarter Algorithms in Detection, Estimation, and Navigation” 

Abstract
By deploying deep learning methods as global optimizers that learn algorithms, we can solve long-standing ill-posed problems in joint parameter estimation. Neural networks can also produce human-interpretable results that de-mystify the algorithm design process, allowing us to rigorously determine hyperparameters such as training lengths. I will introduce two recent works where this methodology is used to tackle open problems in remote sensing and navigation. In underwater environment estimation, I will present a U-Net method for the acoustic mapping of reflective boundaries such as the sea surface and seafloor, with state-of-the-art performance and the new capability of jointly estimating the number of boundaries in the environment. In the field of robotic path planning, I will present a recurrent convolutional neural network (RCNN) method for solving Obstacle Avoiding Rectilinear Steiner Minimum Tree (OARSMT) problems. By learning an algorithm via reinforcement learning, whose intermediate stages are visible to the user, we devise augmentations that yield strong accuracy and runtime performances, which can lead to lower power consumption. Beyond their immediate applications, these new methods point to a general strategy of solving a broad class of joint parameter estimation problems via deep learning.

Biography
Dr. Toros Arikan was born in Mount Kisco, NY, USA, in 1993. He obtained his B.S. and M.S. degrees from the University of Illinois at Urbana-Champaign, with a focus on digital communications and signal processing; and his Ph.D. degree from the Massachusetts Institute of Technology, specializing in localization, tracking, and remote sensing. He is currently a postdoctoral researcher at Rice University, where he continues his research on deep learning methods for environment estimation and algorithm development.

Please join us on Friday, 01/24/25 at 10:20 a.m. in ETB 1020 to learn more and meet Dr. Toros Arikan.

For more on Dr. Toros Arikan visit his website at https://torosarikan.github.io/.

Friday, Nov. 15, 2024
10:20 – 11:10 a.m. (CST)
ETB 1020       

Dr. Ranveer Chandra
General Manager for M365
Chief Technology Officer of Agri-Food at Microsoft

Title: “Data-Driven Agriculture to Sustainably Nourish the World” 

Abstract
Data-driven techniques help boost agricultural productivity by increasing yields, reducing losses and cutting down input costs. However, these techniques have seen sparse adoption owing to high costs of manual data collection and limited connectivity solutions. In this talk we will describe our innovations that leverage Internet of Things, Artificial Intelligence, and Edge Compute in the Farm and Space, to help make affordable digital agriculture solutions. We will also present our product based on this research, which is in preview, and can be used by partners to build their digital agriculture solutions, and how Generative AI can help transform digital agriculture.

Biography
Dr. Ranveer Chandra is the General Manager for M365 and the CTO of Agri-Food at Microsoft. He was previously the Managing Director of Industry Research and led the Networking Research Group at Microsoft Research, Redmond. He was also the Chief Scientist of Microsoft Azure Global. His research has shipped as part of multiple Microsoft products, including VirtualWiFi in Windows 7 onwards, low power Wi-Fi in Windows 8, Energy Profiler in Visual Studio, Software Defined Batteries in Windows 10, and the Wireless Controller Protocol in XBOX One. His research also led to a new product called Azure Data Manager for Agriculture. Ranveer is active in the networking and systems research community and has served as the Program Committee Chair of IEEE DySPAN 2012, ACM MobiCom 2013, and HotNets 2021.

Ranveer has published more than 100 papers and holds over 125 patents granted by the USPTO. His research has been cited by the popular press – such as the Economist, MIT Technology Review, BBC, Scientific American, New York Times, WSJ, among others. He is an IEEE Fellow and an ACM Fellow, and he has won several awards including best paper awards at ACM CoNext 2008, ACM SIGCOMM 2009, IEEE RTSS 2014, USENIX ATC 2015, Runtime Verification 2016 (RV’16), ACM COMPASS 2019, ACM MobiCom 2019, the Microsoft Research Graduate Fellowship, the Microsoft Gold Star Award, the MIT Technology Review’s Top Innovators Under 35, TR35 (2010) and Fellow in Communications, World Technology Network (2012). He was recognized by “Newsweek” as America’s 50 most Disruptive Innovators (2021). Ranveer has an undergraduate degree from IIT Kharagpur, India and a PhD from Cornell University.

 To learn more about Dr. Chandra, visit our webpage at https://www.microsoft.com/en-us/research/people/ranveer/.

Friday, November 8, 2024
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. I-Hong Hou
Professor of CE in ECE
Texas A&M University

Title: “Distributed No-Regret Learning for Multi-Stage Systems with End-to-End Bandit Feedback” 

Abstract
This talk focuses on multi-stage systems with end-to-end bandit feedback. In such systems, each job needs to go through multiple stages, each managed by a different agent, before generating an outcome. Each agent can only control its own action and learn the outcome of the job. It has neither knowledge nor control on actions taken by agents in the next stage. The goal of this paper is to develop distributed online learning algorithms that achieve sublinear regret in adversarial environments.

The setting of this paper significantly expands the traditional multi-armed bandit problem, which considers only one agent and one stage. In addition to the exploration-exploitation dilemma in the traditional multi-armed bandit problem, we show that the consideration of multiple stages introduces a third component, education, where an agent needs to choose its actions to facilitate the learning of agents in the next stage. To solve this newly introduced exploration-exploitation-education trilemma, we propose a simple distributed online learning algorithm, ϵ−EXP3. We theoretically prove that the ϵ−EXP3 algorithm is a no-regret policy that achieves sublinear regret. Simulation results show that the ϵ−EXP3 algorithm significantly outperforms existing no-regret online learning algorithms for the traditional multi-armed bandit problem.

Biography
Dr. Hou is a Professor in the ECE Department of Texas A&M University. His research interests include cloud/edge computing, wireless networks, and machine learning. Dr. Hou received the B.S. in Electrical Engineering from National Taiwan University in 2004 and his M.S. and Ph.D. in Computer Science from the University of Illinois, Urbana-Champaign in 2008 and 2011, respectively. He received the Best Paper Awards in ACM MobiHoc 2020 and ACM MobiHoc 2017, the Best Student Paper Award in WiOpt 2017, and the C.W. Gear Outstanding Graduate Student Award from the University of Illinois at Urbana-Champaign.

 To learn more about Dr. Hou, visit our webpage at https://cesg.tamu.edu/faculty/i-hong-hou/.

Friday, November 1, 2024
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Saurabh Kadekodi
Senior Research Scientist in Storage Analytics
Google

Title: “Data-Driven IO Modeling and Optimizations in Cluster Storage Systems” 

Abstract
In this talk, I will cover two recent data-driven research projects in cluster storage systems — Thesios and Morph. Thesios is a methodology to accurately synthesize full-resolution representative IO traces, and counterfactual “what-if” traces by carefully combining down-sampled IO traces collected from multiple disks attached to multiple storage servers. Representative traces help inform the design and configuration of storage systems on real-world workloads, and counterfactual traces help assess the impact of anticipated or hypothetical new storage policies or hardware prior to deployment. I will also discuss the usefulness of Thesios for academia in order to obtain real-world traces, and the experience in open-sourcing synthesized traces comprising 2.5 billion IO requests.

Morph is the data-driven redundancy adaptation of files stored in cluster storage systems, over their lifetimes to address changes in data temperature and latency requirements. For newly ingested data, commonly stored via 3-way replication, Morph introduces a hybrid redundancy scheme that combines a replica with an erasure coded (EC) stripe, reducing both ingest IO and capacity overheads while enabling free transcode to EC by deleting replicas. For subsequent transcodes to wider, more space-efficient EC configs, Morph exploits Convertible Codes, which minimize data read for EC transcode, and introduces new block placement policies to maximize their effectiveness. Morph is thus designed to optimize redundancy by taking a file-lifetime view and minimizing IO overheads without affecting performance.

Biography
Saurabh Kadekodi obtained his PhD in the Computer Science Department at Carnegie Mellon University (CMU) in 2020 as part of the Parallel Data Laboratory (PDL) under the guidance of Prof. Gregory Ganger and Prof. Rashmi Vinayak. After graduation, Saurabh joined Google as a Visiting Faculty Researcher and is currently a Senior Research Scientist in the Storage Analytics team. Saurabh is broadly interested in designing distributed systems with special focus on the performance and reliability of storage systems.

To learn more about Dr. Kadekodi, visit his homepage at https://www.cs.cmu.edu/~saukad.

Friday, October 18, 2024
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Prabhakar R. Pagilla
Professor, Mechanical Engineering Department
Texas A&M University

Title: “Planning and Control Problems in Robotics for Manufacturing Operations” 

Abstract
There has been a significant growth in the use of articulated robots for automation of manufacturing operations across many industrial sectors, including aerospace, transportation, construction, electronics, etc., with applications ranging from assembly, surface finishing, to material handling and delivery. For example, mechanical surface finishing operations, such as grinding, sanding, polishing, chamfering, etc., are widely employed in many industrial sectors to remove part anomalies and achieve a desired surface finish. Surface finishing has predominantly been a manual operation that is highly labor intensive and requires skilled operators. The key benefits of integrating robots into these environments include consistent surface quality, improved productivity, preventing hazardous exposure to vibration and particulate, significant flexibility for small as well as large batch manufacturing of finished parts, and the potential for re-purposing the robot quickly to adapt to a new part.

One can find many recent research and technology development activities focusing on the technical challenges of integrating robots into these operations in various environments. This talk will provide an overview of some of the essential elements, challenges, and recent advances in integrating robots to improve manufacturing operations, including registration of the workpiece in the robot workspace, path planning and control. A portion of the talk will also discuss some recent work on the benefits and challenges of V2V communication in connected and autonomous vehicles.

Biography
Prabhakar R. Pagilla is a Professor in the Mechanical Engineering Department at Texas A&M University. His formal background and research interests are in dynamic systems and control with applications in robotics, manufacturing, and autonomy. His current research focuses on robot motion planning and control problems in manufacturing, modeling and control of transport behavior of flexible materials in roll-to-roll manufacturing, and cooperative adaptive cruise control systems in connected and autonomous vehicles. He teaches undergraduate and graduate courses in dynamic systems, control, and robotics.

Please join us on Friday, 10/18/24 at 10:20 a.m. in ETB 1020 to learn more and meet Dr. Pagilla.

For more on Dr. Pagilla visit his website at https://pagilla.engr.tamu.edu.

Joseph Jean Boutros
TAMU Qatar
Professor of Electrical & Computer Engineering

Thursday, October 10, 2024
&
Tuesday, October 15, 2024

11:10 a.m. – 12:25 p.m.
@ the WEB 333’s Fishbowl

Title: “Algebra of Codes, Goppa Codes, and Applications to Post-Quantum Cryptography”

Abstract
We introduce linear codes over finite fields. Then, we study Generalized Reed-Solomon (RS) codes. Goppa codes are defined as subfield subcodes of RS codes. Finally, we describe the McEliece cryptosystem based on Goppa codes for asymmetric post-quantum cryptography.

Biography
Joseph Jean Boutros received the M.S. degree in electrical engineering in 1992 and the Ph.D. degree in 1996, both from Ecole Nationale Superieure des Telecommunications (ENST, Telecom ParisTech), Paris, France. From 1996 to 2006, he was with the Communications and Electronics Department at ENST as an Associate Professor. Also, Dr Boutros was a member of the research unit UMR-5141 of the French National Scientific Research Center (CNRS) in Paris. In July 2007, Doctor Boutros joined Texas A&M University at Qatar (TAMUQ) as a full Professor in the electrical engineering program.

Prof. Boutros’ fields of research are codes on graphs, lattice sphere packings, iterative decoding, joint source-channel coding, compressive sensing, space-time coding, physical-layer security, and physical-layer network coding.

Friday, October 11, 2024
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Nina Taft
Principal Scientist/Director at Google

Title: “Leveraging Deep Learning to Understand Users’ Views about Privacy” 

Abstract
Privacy nudges can offer developers suggestions to improve the privacy of their apps.  We design a multi-stage methodology that leverages recent advances in NLP and LLMs to automatically extract privacy insights from smartphone app reviews.  Our analysis pipeline includes a privacy classifier, automated issue tagging for thematic clusters, a classifier to attach emotions to reviews, and extracts temporal and geographic trends.  We apply this methodology to publicly visible app reviews on the Google Play store that span a 10-year period and uncover 12 million instances of privacy-relevant reviews.  We’ll summarize users’ perspectives about smartphone app privacy along multiple dimensions – across a decade of time, from over 200 countries, and across a diversity of app types and privacy topics.  This approach complements traditional user studies by providing developers with actionable feedback from a vast and diverse user base.

Biography
Nina Taft is a Principal Scientist/Director at Google where she leads the Applied Privacy Research group. Nina received her PhD from UC Berkeley and has worked in industrial research labs since then – at SRI, Sprint Labs, Intel Berkeley Labs, and Technicolor Research before joining Google.  For many years, Nina worked in the field of networking, focused on Internet traffic modeling, traffic matrix estimation, and intrusion detection. In 2017, she received the top-10 women in networking IEEE N2Women award. In the last decade, she has been working on privacy enhancing technologies with a focus on applications of machine learning for privacy. She has been chair of the SIGCOMM, IMC and PAM conferences, has published over 100 papers, and holds 10 patents.

Please join us on Friday, 10/11/24 at 10:20 a.m. in ETB 1020!

Friday, October 25, 2024
10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Vijay Narayanan
Pennsylvania State University
Professor of Computer Science & Engineering and Electrical Engineering
Associate Dean of Innovation
Director of the Penn State Center for AI Intelligence Foundations and Engineering Systems

Title: “Designing Emerging Computing Systems with Ferroelectric Devices” 

Abstract
This talk will present a brief overview of advances in ferroelectric devices and their integration into computing systems to provide novel functionality and energy efficiency in various data intensive applications. The talk will emphasize on cross-stack design opportunities in designing stacked intelligent 3D memory systems.

Biography
Vijaykrishnan Narayanan is an Evan Pugh University Professor and Robert Noll A. Chair Professor of Computer Science and Engineering and Electrical Engineering at Pennsylvania State University.
He is a Fellow of ACM, IEEE, AAAS and the National Academy of Inventors. He serves as associate director of DoE 3DFeM center, thrust lead for DARPA/SRC PRISM center, associate Editor-in-Chief of IEEE Micro, the academic coordinator for the India-US Defense Acceleration Ecosystem and Associate Executive Director of AI for the GeoEd Foundation.

To learn more about Dr. Narayanan go to https://sites.psu.edu/vijaykrishnannarayanan.

Join us on Friday, 10/25/24 at 10:20 a.m. in ETB 1020 to hear Dr. Narayanan present in-person. 

Friday, October 4, 2024

10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Li-C. Wang
Professor in Electrical and Computer Engineering
University of California, Santa Barbara

Title: “Harnessing the Power of Large Language Models” 

Abstract
The emergence of Large Language Models (LLMs) has transformed how we apply Machine Learning (ML) in the field of semiconductor test. Recent advancements in LLMs showcase their impressive ability to engage in meaningful dialogue across a wide range of topics, answer complex questions, and even generate code.

In this talk, I will share our experience in harnessing the power of LLMs to develop an AI agent specifically for semiconductor test data analytics. Our approach centers on the integration of a Knowledge Graph (KG) advocate for an end-to-end methodology that positions the KG as a critical component. I will introduce a novel paradigm, Decision-Support ML (DSML), and explain its implementation in common test data analytics workflows. Using wafermap analytics as a case study, we have demonstrated how we built IEA-Plot, an LLM-assisted AI solution, by leveraging typical LLM functionalities. I will show the real-world application of IEA-Plot on test data collected from a recent production line.

Biography
Li-C. Wang is a professor in the ECE department at the University of CA, Santa Barbara. He received his PhD in 1996 from the University of Texas at Austin and was previously with Motorola PowerPC Design Center.

Starting from 2003, his research has focused on investigating how machine learning could be utilized in design and test flows, where he had published more than 100 papers and supervised 22 PhD thesis on the related subjects. Prior to that, his research spanned across multiple topics in EDA and test including microprocessor test and verification, statistical timing analysis, defect-oriented testing, and SAT solvers. He received 10 Best Paper Awards and 2 Honorable Mentions paper awards from major conferences including recent best paper awards from ITC 2022, ITC 2020, VTS 2016, and VLSI-DAT 2019. He is the recipient of the 2010 Technical Excellence Award from Semiconductor Research Corporation (SRC) for his research contributions in data mining for test and validation. He is the recipient of the 2017 IEEE-TTTC Bob Madge Innovation Award. He is an IEEE fellow and served as the General Chair of the International Test Conference (ITC) in 2017, 2018, and 2023.

To learn more about Dr. Wang go to https://iea.ece.ucsb.edu/.

Join us on Friday, 10/4/24 at 10:20 a.m. in ETB 1020 to learn more about Dr. Wang’s approach!

Friday, September 20, 2024

10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Haris Pozids
Scientist & Manager of Infrastructure AIOPS
IBM Research Europe – Zurich, Switzerland

Title: “Using Generative AI Technology to Transform Customer Support Services” 

Abstract
Agent Assist is a software tool, that enables remote technical support agents solve clients’ problems faster and helps newly hired support agents increase their productivity and reduce the time required to learn and resolve cases. Agent Assist integrates tightly with the IBM Cognitive Support Platform (CSP), the tool that IBM support agents use to troubleshoot customer cases. It receives a real-time notification from CSP every time a new case is opened and pushes back to CSP recommendations for resolving that case, based on analysis of past, resolved, similar cases and relevant product documentation. Agent Assist leverages watsonx.ai for retrieving similar historical cases and for past case resolution extraction and summarization, using generative large language models (LLMs). It achieves fast response of less than 60 seconds end-to-end by performing most time-consuming LLM and indexing operations offline.

The transformation of customer support service into a more automated and efficient business by the use of AI is strategic for many businesses, promising higher efficiency, lower cost of operation and improved customer satisfaction. The Agent Assist technology is enabling and accelerating this transformation by harnessing the power of generative AI and by scaling across multiple Support products and functions.

Biography
Haris Pozidis manages the AI for Infrastructure group at IBM Research in Zurich, Switzerland, which focuses on the design of algorithms for scalable and accelerated machine learning, on the development of Flash memory controllers and Computational Storage, and on AI-infused systems for improving cloud resiliency and operations.

Dr. Pozidis holds over 160 US patents in areas ranging from scalable machine learning systems to storage systems and solid-state memory technology. He has co-authored more than 120 journal and conference publications in the above areas. He is a Senior Member of the IEEE and a Principal Research Scientist at IBM Research.

To learn more about Dr. Pozids go to https://research.ibm.com/people/haris-pozidis.

You are welcome to join us on Friday, 9/20/24 at 10:20 a.m. in ETB 1020!

Friday, September 13. 2024

10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Wen-mei W. Hwu
Senior Director of Research at NVIDIA
Professor Emeritus l Electrical & Computer Engineering
University of Illinois, Urbana-Champaign

Title: “BaM: System Architecture and Software Stack for Accelerating Compute-Directed Access to Massive Datasets”

Abstract
Compute Devices have traditionally relied on OS services to bring data into the memory in bulk before performing algorithmic computation on the individual data-structure elements. For example, Graphics Processing Units (GPUs) have relied on the host CPU services to bring chunks of storage data into its device memory for use by compute kernels. This approach is well-suited for GPU applications with known data access patterns that enable partitioning of their dataset to be processed in a pipelined fashion in the GPU. However, emerging applications such as graph and data analytics, recommender systems, and graph neural networks, require fine-grained, data-dependent and sparse access to vast feature vectors and embedding datasets. CPU services are unsuitable for these applications due to high CPU-GPU synchronization overheads, I/O traffic amplification, and low CPU software throughput. GPU-initiated access avoids these overheads by removing the CPU from the storage control path and, thus, can potentially support these applications at much higher speed. However, there is a lack of systems architecture and software stack that enable efficient GPU-initiated storage access for applications today. I will present a vision for enabling fast, compute-directed sparse access to massive datasets, the BaM system architecture to realize this vision, and the BaM software stack that efficiently supports emerging applications on existing and upcoming GPUs.

Biography
Wen-mei W. Hwu is a Senior Distinguished Research Scientist and Senior Director of Research at NVIDIA. He is also a Professor Emeritus and the Sanders-AMD Endowed Chair Emeritus of ECE at the University of Illinois at Urbana-Champaign. His research is in the architecture, algorithms, and infrastructure software for data intensive and computational intelligence applications. He served as the Illinois director of the IBM-Illinois Center for Cognitive Computing Systems Research Center (c3sr.com) from 2016 to 2020. He was a PI of the NSF Blue Waters supercomputer project. He received the ACM SigArch Eckert-Mauchly Award, ACM/IEEE Maurice Wilkes Award, the ACM Grace Murray Hopper Award, the IEEE Computer Society Charles Babbage Award, the ISCA Influential Paper Award, the MICRO Test-of-Time Award, the IEEE Computer Society B. R. Rau Award, the CGO Test-of-Time Award, numerous best paper awards, numerous teaching awards, and the Distinguished Alumni Award in CS of the University of California, Berkeley. He is a Fellow of IEEE and ACM.

To learn more about Dr. Hwu: https://research.nvidia.com/person/wen-mei-hwu

Please join us on Friday, 9/13/24 at 10:20 a.m. in ETB 1020.

Friday, September 6, 2024

10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Srinivas Shakkottai
Professor, Electrical & Computer Engineering, Texas A&M University                  

Title: “Structured Reinforcement Learning in NextG Cellular Networks”

Abstract
NextG cellular networks face increasing demands for intelligent control, especially with the advent of softwarized Open Radio Access Networks (O-RAN) and diverse user applications. We present EdgeRIC, a real-time RAN Intelligent Controller (RIC) co-located with the Distributed Unit (DU) in the O-RAN architecture, enabling sub-millisecond AI-optimized decision-making. We propose a constrained reinforcement learning (CRL) approach for developing such real-time strategies, showing that these algorithms can be trained with only a logarithmic increase in complexity compared to traditional RL. We introduce structured learning using threshold and Whittle index-based policies, which provides low-complexity learning and scalable, real-time inference for optimizing resource allocation and enhancing user experience. For media streaming, we prove the optimality of a threshold policy and develop a soft-threshold natural policy gradient (NPG) algorithm that prioritizes clients based on video buffer length, achieving inference times of about 10μs and improving user quality of experience by over 30%. Additionally, we leverage Whittle indexability to simplify resource allocation, ensuring service guarantees such as ultra-low latency or high throughput by training neural networks to compute constrained Whittle indices. Our Whittle index approach, implemented on EdgeRIC, achieves allocation decisions within 20μs per user, enhancing service guarantees across standardized 3GPP service classes, making a case for structured, scalable reinforcement learning for real-time control of NextG networks.

Biography
Srinivas Shakkottai received his PhD in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign in 2007, after which he was a postdoctoral scholar in Management Science and Engineering at Stanford University.  He joined Texas A&M University in 2008, where he is currently a professor at the Dept. of Electrical and Computer Engineering and at the Dept. of Computer Science and Engineering (by courtesy).  His research interests include multi-agent learning and game theory, reinforcement learning, communication and information networks, networked markets, as well as data collection and analytics.  He co-directs the Learning and Emerging Network Systems (LENS) Laboratory and the RELLIS Spectrum Innovation Laboratory (RSIL).  He has served as an Associate editor of IEEE/ACM Transactions on Networking and the IEEE Transactions on Wireless Communications.  Srinivas is the recipient of the Defense Threat Reduction Agency (DTRA) Young Investigator Award and the NSF CAREER Award, as well as research awards from Cisco and Google. His work has received honors at fora such as ACM MobiHoc, ACM eEnergy and the International Conference on Learning Representations.  He has also received an Outstanding Professor Award, the Select Young Faculty Fellowship, and the Engineering Genesis Award (twice) at Texas A&M University.

For more on Dr. Shakkottai and his work, go to Srinivas Shakkottai (tamu.edu).

Please join us on Friday, 9/6/24 at 10:20 a.m. in ETB 1020.

Friday, August 31, 2024

10:20 – 11:10 a.m.  (CST)
ETB 1020           

Dr. Bobak Mortazavi
Associate Professor, Computer Science & Engineering, Texas A&M University                  

Title: “Sensing and Modeling for Personalized Cardiovascular Digital Health”

Abstract
Recent sensing and modeling techniques have led to advancements in clinical modeling, resulting in advancements in interpretation, evaluation, and risk prediction in cardiovascular outcomes. Large datasets, a focus on integrating electronic health records and clinical trials data has enabled improvements in cardiovascular care. However, there still remain significant gaps in patient care in personalized and remote setting in both diagnosis and recovery. In this talk we explore sensing and modeling for remote health for personalized cardiovascular care, moving towards continuous monitoring of clinical biomarker signals in remote settings for clinical diagnosis.

Biography
Bobak Mortazavi, PhD, is an Associate Professor of Computer Science & Engineering and part of the Center for Remote Health Technologies and Systems at Texas A&M University and holds an affiliation with the Yale University School of Medicine’s Center for Outcomes Research and Evaluation. His research focuses on the intersection of wearable technology, machine learning, and cardiovascular-focused clinical outcomes research, to develop longitudinal, personalized models of health. He has made important contributions in enabling wearable sensing technologies for personal health monitoring and integrating machine learning modeling for improving the use of this data in the context of clinical outcomes, with his work supported by awards from DARPA, the NSF, and the NIH.

For more on Dr. Mortazavi and his work, go to https://engineering.tamu.edu/cse/profiles/mortazavi-bobak.html) & STMI Lab – Home

Please join us on Friday, 8/31/24 at 10:20 a.m. in ETB 1020.

Friday, August 23, 2024
10:20 a.m.  (CST)
ETB 1020           

Dr. Srikanth Saripalli
Professor, Department of Mechanical Engineering, Texas A&M University                  

Title: “Autonomy in the Wild: Perception and Control for Off-Road Autonomous Vehicles”

Abstract
This talk focuses on perception and planning algorithms for autonomous vehicles in off-road situations. A particular emphasis is on why off-road vehicles are different than on-road vehicles and how can we solve autonomy in the off-road domain. A major portion of the talk will be on applications of the above algorithms to real vehicles and the lessons that we have learned i.e. what worked and what didn’t and how we should go about building such systems.

Biography
Srikanth Saripalli is a Professor in the Mechanical Engineering Department and the Director for Center for Autonomous Vehicles and Sensor Systems (CANVASS) at Texas A&M University. He holds the J. Mike Walker ’66 Professorship. His research focuses on robotic systems: particularly in air, water and ground vehicles and necessary foundations in perception, planning, control and system integration for this domain. He is currently leading several efforts in off-road autonomous ground vehicles. He has also led several long-term (> 6 month) on-road deployments of autonomous 18-wheeler trucks and slow-moving shuttles in Texas. He is currently interested in developing and deploying Autonomous Shuttles on campus and in cities. He is also interested in developing such autonomous shuttles for mobility challenged and para transit applications.

For more on Dr. Saripalli and his work, go to https://unmanned.tamu.edu.

Please join us on Friday, 8/23/24 at 10:20 a.m. in ETB 1020.

Monday, May 13, 2024

10:15 – 11:15 a.m.  (CST)
WEB 236C

Dr. Naehyuck Chang
Executive Vice President | Samsung SDI America

Title: “The Challenges and Opportunities in the Mobility Electrification“

Abstract
The electrification of mobilities is essential for sustainability, and governments are pushing hard to expedite electric vehicle penetration. As a result, many electric vehicles on the road are present today. Nevertheless, the battery industry faces significant challenges starting in late 2023, which will continue for years. In this talk, we will introduce industry perspective challenges and opportunities for mobility electrification in both the technical and business aspects. The technical aspects to be covered are automotive battery requirements such as energy density, charging time, lifetime, cost, and safety. We will talk about the infrastructure issues for electric vehicle charging. As for the business aspects, we will discuss the vehicle electrification roadmap, government support, marketing challenges, cost demands, battery raw material costs, etc. This talk is the first to introduce battery technical evolution in the context of “Battery Technology Scaling,” analogous to semiconductor technology scaling. We will also summarize the expected academic contributions to the electrification of mobilities.

Biography
Naehyuck Chang is an Executive Vice President at Samsung SDI America. He was the Head of Development at Samsung SDI Headquarters from 2021 to 2023. Dr. Chang was in charge of all automotive and energy-storage product developments, from cells to systems. He is the Founder of EMVcon, Inc., Irvine, CA, a vehicle electrification company funded by Samsung. Dr. Chang’s research interests include low-power computing, cyber-physical systems, and Design Automation of Things, such as systematic design and optimization of mobility electrification, energy storage systems, and energy harvesting. From 1997 to 2014, Dr. Chang was a professor at the Department of Computer Science and Engineering at Seoul National University. Since 2014, he has been a professor at the Department of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea. Dr. Chang is a Fellow of the ACM and a Fellow of IEEE for his contributions to low-power design and computing. Dr. Chang is a member of the National Academy of Engineering of Korea. Dr. Chang was the Chair and the Past Chair of the ACM Special Interest Group on Design Automation. Dr. Chang was the TPC Co-Chair of DAC 2016, ASP-DAC 2015, ICCD 2014, CODES+ISSS 2012, and ISLPED 2009, and the General Co-Chair of VLSI-SoC 2015, ICCD 2015 and 2014, and ISLPED 2011. Dr. Chang was the Editor-in-Chief of the ACM Transactions on Design Automation of Electronics Systems.

Please join on Monday, 5/13/24 at 10:15 a.m. in WEB 236C