Plenary Speakers

Monday | May 26, 2025 | 9:30 - 10:30

Imagine a silicon chip that can seamlessly adapt to any environment, excel across different tasks, and operate with limited energy—all without exhausting its memory. This capability, known as continual learning, enables the chip to learn continuously from evolving data, while excelling at past, current, and future skills simultaneously. 

Although such adaptability is quite natural for biological agents, such as humans, achieving it in traditional machine learning models has proven quite challenging. Biological brains appear to have evolved to learn continually through their lifetimes, drawing power as low as ∼ 0.1 watts of Adenosine 5’-Triphosphate (ATP) for cortical computation.

Monday | May 26, 2025 | 15:00 - 16:00

Quantum computing is a new paradigm that exploits the fundamental principles of quantum mechanics, such as superposition and entanglement, to tackle problems in mathematics, chemistry and material science that are well beyond the reach of supercomputers. Despite the intensive worldwide race to build a useful quantum computer, it is conservatively projected to take decades before reaching the state of useful quantum supremacy. The main challenge is that qubits operate at the atomic level, thus are extremely fragile, and difficult to control and read out. The current commercial state-of-the-art implements a few dozen superconducting qubits in a highly specialized technology and cools them down to a few tens of millikelvin. The high cost of cryogenic cooling and bulky wiring prevents its widespread use. A companion classical electronic controller, needed to control and read out the qubits, is mostly realized with room-temperature laboratory instrumentation. The recent progress with integrating the latter on a CMOS chip and placing it at the 4-kelvin stage within the same cryogenic chamber has made a tremendous progress towards the total system integration, but still it does not even start addressing the basic question of how to scale up to the thousands or millions of qubits needed for practical quantum algorithms. 

Tuesday | May 27, 2025 | 9:30 - 10:30

Large Language Models (LLMs) have demonstrated exceptional performance across numerous generative AI applications, but require large model parameter sizes. These parameters range from several billion to trillions, leading to significant computational demands for both AI training and inference. The growth rate of these computational requirements significantly outpaces advancements in semiconductor process technology. Consequently, innovative IC and system design techniques are essential to address challenges related to computing power, memory, bandwidth, energy consumption, and thermal management to meet AI computing needs.

In this talk, we will explore the evolution of LLMs in the generative AI era and their influence on AI computing design trends. For AI computing in data centers, both scale-up and scale-out strategies are employed to deliver the huge computational power required by LLMs. Conversely, even smaller LLM models for edge devices demand more resources than previous generations without LLMs. Moreover, edge devices may also act as orchestrators in device-cloud collaboration. These emerging trends will significantly shape the design of future computing architectures and influence the advancement of circuit and system designs.

Tuesday | May 27, 2025 | 14:00 - 15:00

As the demand for energy-efficient connectivity, embedded memories and better sensing capabilities grows, ultra-low power CMOS technologies such as 22FDX and FinFETs are becoming essential for advancing applications in Edge AI, Power management, and Quantum computing.

This talk will explore the latest innovations in design and how they are enabling the deployment of sophisticated AI solutions at the edge, when combined with GlobalFoundries's ultra-low power CMOS technologies.

It will focus on Edge AI Automotive sensors with ADAS 77GHz Radar sensor as an example and how fast transistors, low loss BEOL, power efficiency and low noise are critical to realizing this integrated radar solution for fully autonomous cars and imaging radars.

Wednesday | May 28, 2025 | 9:30 - 10:30

The evolution of satellite communication systems is increasingly driven by the need for high-performance, flexible, and efficient solutions, particularly in the design and implementation of circuits and systems for next-generation satellite payloads. Software defined payloads leveraging on On-Board Digital Signal Processing (OBP), Multi-Beam Antennas (MBAs) and Beam-Forming Networks (BFNs) are pivotal technologies in addressing the challenges posed by emerging communication standards, such as 5G/6G, as well as advanced satellite communication and remote sensing systems.