Plenary Speakers

Abstract
Abstract
Amount of compute power required in today’s SoCs, especially in AI applications, is outpacing Moore’s Law by a wide margin. Orders of magnitude compute are needed to keep pace with this new era of scaling and system complexity, otherwise known as the SysMoore era. In this presentation we examine the challenges driving the next wave of innovative design solutions. Starting with architecture exploration to silicon-lifecycle management, these solutions will help engineers create designs that keep pace with the compute power SoC needs of today and tomorrow.

Abstract
The semiconductor industry is at the beginning of a new renaissance – one that will change the world, yet again. The era of specialized chips and software workloads has arrived, and with that comes new opportunities to digitally transform many industries- from the core to the edge. In this talk, we will look at the worldwide economic landscape and how semiconductors are leading the way to digitally transform many industries- from factories to cars to shopping. We will see why specialization is central to the next wave of growth and show how this is bringing dramatic changes in design, verification, and manufacturing of SoCs. Whether it is SoCs at the core (datacenter) or at the edge (imaging sensors), new types of computing are emerging. With new demands on computing, new computer architectures are emerging in SoC designs to enable power-efficient information processing. The verification of these ICs in the systems they operate is rapidly becoming the biggest challenge in getting these products to market and then monitoring their operation in the system during the lifetime of the SoC. We will examine what is happening to this world of new SoCs in this new semiconductor super-cycle.

Abstract
In today’s global environment, everything is about data…. Big data. Our economy is, in fact, driven by all aspects of data handling (creation, transmission, storage and analytics), and high performance processing is at the heart of each of these aspects. We are experiencing the need for advances in data handling and processing as never before. Modernization of data creation and handling has heightened requirements and stimulated advancements in the areas of AI, Machine Learning and data transmission. The high performance computing demands have increased nonlinearly to meet the increasing needs at data origination, edge computing, transmission backbones and datacenters. The design objective metrics have not really changed throughout these evolutions in computing, but the emphasis has shifted. Where the focus was previously only on frequency, and then on isolated low power-centric solutions, the emphasis is now on the total energy: Performance per watt is everything. In addition to Energy efficiency (performance/watt), productivity and capacity is extremely important across a wide variety of compute areas. The demands are present in AI/ML, networking, cloud computing, and autonomous/advanced driver assistance systems.
The challenges require not only advances in process technology and associated tools and flows, but also a tight collaboration with the eco system and design community. Through proactive industry collaboration Cadence has enhanced the high performance computing platform to address and manage performance, power and productivity for the full variety of applications. We will provide a preview to Cadence’s wholistic solution addressing industry needs through this session.