(The talk will be given in English)

Speaker:     Dr. Ofer Shacham

                    Co-founder and CEO of Majestic Labs

Wednesday, January 29^th, 2025

15:00 - 16:00

The era of domain specific architectures -- Case in point: artificial intelligence

Abstract

The end of Dennard scaling marks the dawn of an era where all designs are power constrained. An era where the only way to increase performance is to improve power efficiency. This is the era of domain specific architectures. In this talk, our focus is on specialized extremely efficient computing for application in the most impactful domain of the century—artificial intelligence. Furthermore, artificial intelligence is the dominant driver for demand for compute in both mobile and data center use cases. In this talk we will see how domain specific architectures deliver the exponential performance growth we so desperately desire in order to feed the AI beast. 

Short Bio

Dr. Ofer Shacham is a co-founder and CEO of Majestic Labs, a US-Israeli startup, rearchitecting AI server systems to topple the Memory Wall. He was most recently VP and Head of Silicon at Meta Reality Labs creating the Meta AR glasses compute system known as Orion. Prior to Meta, Dr. Shacham worked for Google where he led Silicon Design and Implementation for the Consumer Hardware organization and developed Google’s Pixel Visual Core and Pixel Neural Cores for AI, some generations of the Titan security chip and early phases of the Pixel 6 Tensor Chip. Dr. Shacham joined Google through the acquisition of his startup, Chip Genesis, which helped organizations create specialized extremely efficient chips. Prior to Stanford, Dr. Shacham worked for IBM R&D labs in Israel. Shacham also served in an elite Israeli Navy Unit. Dr. Shacham holds a double major BSc in EECS from Tel Aviv University and a MSc, Ph.D., and Postdoc degrees from Stanford University, where he later served as Academic Staff. Dr. Shacham enjoys paying it forward by teaching VLSI and Computer Architecture at Tel Aviv University.

השתתפות בסמינר תיתן קרדיט שמיעה לתלמידי תואר שני ושלישי = עפ"י רישום שם מלא + מספר ת.ז. בטופס הנוכחות שיועבר באולם במהלך הסמינר

(The talk will be given in English)

Speaker:     Dr. Ohad Trabelsi

Computer Science at Toyota Technological Institute at Chicago

Sunday, January 26^th, 2025

15:00 - 16:00

New Developments in Fine-Grained Complexity

Abstract

Fine-grained complexity has emerged in the last decade as a powerful tool for understanding the computational limits of problems in P. By linking a few key hardness assumptions to a broad range of problems, this tool provides a systematic approach to addressing longstanding challenges in algorithm design. In this lecture, I will explore fine-grained complexity, primarily as a guide to faster algorithms—highlighting recent advances in the Gomory-Hu problem, along with progress on a few other related topics.

Short Bio

I am a Research Assistant Professor in Computer Science at Toyota Technological Institute at Chicago. Previously, I was a postdoctoral fellow in Computer Science at The University of Michigan. I obtained a Ph.D. degree at Weizmann Institute, where I was fortunate to be advised by Prof. Robert Krauthgamer. My main research interests lie at the intersection of fine-grained complexity and algorithm design.

השתתפות בסמינר תיתן קרדיט שמיעה לתלמידי תואר שני ושלישי = עפ"י רישום שם מלא + מספר ת.ז. בטופס הנוכחות שיועבר באולם במהלך הסמינר

ٍاخ

  -סמינר זה יחשב כסמינר שמיעה לתלמידי תואר שני ושלישי-  This Seminar Is Considered A Hearing Seminar For Msc/Phd Students