Keynote Speakers
Albert Greenberg
Uber, USA
Albert Greenberg is the Vice President of Platform Engineering at Uber, where he leads the teams responsible for the company's data center infrastructure, networking, and compute platforms. Prior to joining Uber, he was a Technical Fellow and Corporate Vice President at Microsoft, creating the networking systems that drive Microsoft Azure. A recognized pioneer in software-defined networking (SDN) and data center networks, Albert is a Fellow of the ACM, IEEE, and a member of the National Academy of Engineering. He holds a Ph.D. in Computer Science from the University of Washington.
Alexandre Proutiere
KTH, Royal Institute of Technology, Sweden
Statistical and Computational Trade-offs in Network Optimization
Abstract:
This talk presents recent advances in solving combinatorial online network optimization problems, with a focus on understanding the statistical and computational trade-offs involved. We examine both the sample complexity and the regret associated with these problems. We introduce a learning algorithm that is both statistically optimal and computationally efficient, running in polynomial time. We further develop specialized algorithms for scenarios where the network-wide utility can be modeled as a factor graph. The performance of our methods is demonstrated through numerical experiments, including synthetic benchmarks and real-world applications to optimizing antenna tilt coordination in radio communication networks.
This talk presents recent advances in solving combinatorial online network optimization problems, with a focus on understanding the statistical and computational trade-offs involved. We examine both the sample complexity and the regret associated with these problems. We introduce a learning algorithm that is both statistically optimal and computationally efficient, running in polynomial time. We further develop specialized algorithms for scenarios where the network-wide utility can be modeled as a factor graph. The performance of our methods is demonstrated through numerical experiments, including synthetic benchmarks and real-world applications to optimizing antenna tilt coordination in radio communication networks.
Alexandre Proutiere is professor at the Decision and Control Systems division in the EECS school at KTH, Royal Institute of Technology (Stockholm, Sweden). His research interests include reinforcement learning, unsupervised learning, statistics, stochastic processes, and control; with applications to communication and social networks, and cybersecurity. Before joining KTH, he was a researcher at Microsoft Research (Cambridge, UK) from 2007 to 2011 and at France Telecom R&D from 2000 to 2006. He has also been an invited lecturer and researcher at the computer science department Ecole Normale Superieure (Paris) from 2004 to 2006. He received a PhD in mathematics from Ecole Polytechnique, graduated in mathematics from Ecole Normale Superieure (Paris), and has an engineering degree from Telecom Paris. He was the recipient of an ERC consolidator grant, and in 2009, received the ACM Sigmetrics rising star award.
Ranjita Bhagwan
Google, India
Addressing Challenges in Observability of Google's Autonomous Network
Abstract:
Google owns and operates one of the world's largest networks, supporting billions of users. Today, this network not only supports the users of Google's various applications such as Gemini, Search, Youtube, Gmail and Maps, it also forms a critical part of the infrastructure supporting enterprise customers of the Google Cloud Platform. The scale and complexity of such varied applications sets the stage for autonomous operations, which require continuous observation of the performance and reliablity of the network and acting upon the observations in a timely way. In this talk, I will present some of the challenges in autonomous network operations and network observability, and how we are addressing them.
Google owns and operates one of the world's largest networks, supporting billions of users. Today, this network not only supports the users of Google's various applications such as Gemini, Search, Youtube, Gmail and Maps, it also forms a critical part of the infrastructure supporting enterprise customers of the Google Cloud Platform. The scale and complexity of such varied applications sets the stage for autonomous operations, which require continuous observation of the performance and reliablity of the network and acting upon the observations in a timely way. In this talk, I will present some of the challenges in autonomous network operations and network observability, and how we are addressing them.
Ranjita Bhagwan is currently a Principal Engineer at Google, working on making Google's network highly reliable. Recently, her work has focused on using data-driven approaches to improve networks and systems. She is an ACM Distinguished Member, INAE Fellow, and is the recipient of the 2020 ACM India Outstanding Contributions to Computing by a Woman Award. She received her PhD and MS in Computer Engineering from University of California, San Diego and a BTech in Computer Science and Engineering from the Indian Institute of Technology, Kharagpur.
Stefano Salsano
University of Rome Tor Vergata, Rome, Italy
Abstract
Abstract:
The talk will focus on the “AI fabric.” The participants will learn the architectures of current AI fabrics (i.e., data center infrastructure) for connecting tens or hundreds of GPUs with 400G or 200G links, The current developments proposed by the vendors also within the Ultra Ethernet consortium will be reviewed. After this overview, the application of SRv6 to an AI fabric will be discussed. The solution based on SRv6 will be analyzed and compared with the state-of-the-art solutions.
The talk will focus on the “AI fabric.” The participants will learn the architectures of current AI fabrics (i.e., data center infrastructure) for connecting tens or hundreds of GPUs with 400G or 200G links, The current developments proposed by the vendors also within the Ultra Ethernet consortium will be reviewed. After this overview, the application of SRv6 to an AI fabric will be discussed. The solution based on SRv6 will be analyzed and compared with the state-of-the-art solutions.
Stefano Salsano is a Full Professor in the Electronic Engineering Department of the University of Rome Tor Vergata. He received his Laurea degree in 1994 (Univ. of Rome Tor Vergata) and his Ph.D. in 1998 (Univ. of Rome “La Sapienza”). From 2018 to 2022 he has been the Coordinator of the Master’s Degree in “ICT and Internet Engineering” and of the Bachelor Degree “Ingegneria di Internet”. He is currently the president of the ICT regulation authority of San Marino Republic. He participated in 19 research projects funded by the EU and the ESA, being a Work Package leader or unit coordinator in 8, technical coordinator in two (Simple Mobile Services, SCISSOR), and project coordinator in one (Superfluidity). He has been the principal investigator in several research and technology transfer contracts funded by industries. He has led the development of several testbeds and demonstrators in the context of EU projects, most of them released as Open Source software. His research interests include Networking for AI, Network Function Virtualization, Software Defined Networking, and Cybersecurity. He co-authors an IETF RFC and more than 180 papers and book chapters. He has been the General Chair of the 14th International Conference on Network and Service Management (CNSM 2018). Since 2017, he has been cooperating with Cisco Systems on Segment Routing and received four Research Grants from Cisco Systems.
.png)
