Inside the AI Fabric: Scaling the Networks Behind Tomorrow's AI Factories
Already registered? Click here to log in.
A certificate of attendance will be offered.
Summary
Artificial intelligence is fundamentally changing how data centers are designed. As AI factories evolve from clusters of thousands of GPUs to infrastructures supporting hundreds of thousands of accelerators, the network fabric has become one of the defining constraints on performance, scalability, power efficiency, and cost. Building these environments requires more than faster optics. Operators must balance scale-up architectures within AI clusters, scale-out networking across massive GPU fabrics, and increasingly scale-across strategies that connect distributed training environments into a unified AI infrastructure. At the same time, technologies including co-packaged optics (CPO), linear pluggable optics (LPO), silicon photonics, optical engines, advanced packaging, and next-generation Ethernet are rapidly reshaping what's possible inside the data center. Our panelists explore which technologies are ready for deployment, where the industry's biggest bottlenecks remain, how networks are validated and operated at unprecedented scale, and which architectural decisions organizations should make now to prepare for the next generation of AI infrastructure. Attendees will gain practical insight into the optical and networking innovations enabling tomorrow's AI factories, while hearing firsthand how operators are balancing immediate deployment requirements with long-term architectural roadmaps.
Panelists
Christian Urricariet
Senior Director, Product and Strategic Marketing
Lumentum
Christian Urricariet is Senior Director of Product and Strategic Marketing at Lumentum, with over 25 years of experience in photonics and optical networking, primarily in product management roles. Across his career, his work has focused on bringing advanced optical interconnect technologies from development to market, with a current focus on AI and hyperscale cloud data centers and large-scale network architectures. Prior to Lumentum, he spent 21 years at Finisar and 3.5 years at Intel, contributing to multiple generations of optical connectivity solutions for cloud and enterprise data centers, as well as telecom networks. He holds a degree in Electrical Engineering with a specialization in optoelectronics
Rishi Chugh
Vice President and General managers, Cloud Switch Business,
Marvell
Rishi Chugh is VP/GM of the Cloud Switch Business at Marvell, where he leads the switch product line serving the Cloud and AI market segment. He brings over 20 years of experience in product management and the design of infrastructure ASICs and high-speed interconnect technologies, and is currently driving the enablement of CPO, NPO, and CPC technologies for scale-up and scale-out fabric deployments. Prior to Marvell, Mr. Chugh held senior roles at Cadence, Cavium, Broadcom, Altera, LSI Logic and Artisan Components (acquired by ARM). He holds a Bachelor’s degree in Electronics from the University of Bombay, a Master of Science in Electrical Engineering from San Jose State University and an MBA from Santa Clara University.
Neel Patel
General Manager Optical Component Technology
Nokia
Neel Patel is 30 year veteran of the semiconductor industry. He is currently General Manager for Nokia Optical Component Technology, responsible for a range of internally developed solutions, including InP and Silicon Photonics PICs for inter and intra datacenter, CW lasers, coherent optical engines, DSP, EIC drivers, and the like. He also serves as Applications Engineering leader for components and coherent pluggable solutions. Prior to joining the Infinera/Nokia He worked at Cisco as a hyper scaler focused sales leader, and helped lead and start the sustained business of Cisco serving hyperscalers with data center component, white box, and other disaggregated solutions. He earned his BSEE from University California, Irvine.
