Data center hardware infrastructure is undergoing an unprecedented period of expansion, in large part driven by the rapidly growing computational demands of artificial intelligence (AI). Large and complex compute clusters consisting of tens of thousands of processors are today required to meet the needs of generative AI. Due to increasing AI model complexity and the growing use of AI in end-user applications, the demands on data center infrastructure are accelerating. The growth in infrastructure needed for AI is driving a very rapid growth in data center power consumption.

In this webinar hosted by the Optical Communications Technical Group, Erman Timurdogan from Lumentum will discuss the network power consumption of large-scale AI training implementations and the impact of several emerging approaches to increase power efficiency. The potential of certain key optical innovations will be discussed to reduce network power consumption by up to 80% for AI large language model training.

Subject Matter Level: Intermediate - Assumes basic knowledge of the topic

What You Will Learn:
• AI networks
• Next generation optics
• Optical transceivers

Who Should Attend:
• Researchers and Industry members at any career stage, wanting to learn more about Optical transceivers and AI networks

About the Presenter: Erman Timurdogan from Lumentum

Erman Timurdogan, Ph.D., is the director of siliconphotonics R&D at Lumentum for next-generation integrated laser-on-silicon and transceiver silicon photonics platforms. Prior to joining Lumentum, he was the director of photonics high-speed design at Rockley Photonics, and previously served as director of optical communications and process design kit engineering at Analog Photonics. Timurdogan received his doctorate in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology (MIT). He contributed to numerous books, patents, and publications in the field of silicon photonics, https://scholar.google.com/citations?user=5oxSntMAAAAJ&hl.