Part 1: How AI Drives the Ongoing Datacenter Transformation - Adrien Sanchez 

• The growth of AI workload in datacenter, both for AI training and inferences
• AI training and inference datacenter workloads differences and current trends for each
• Processor technologies that are used to handle these workloads, and current evolutions. The ongoing impact of LLM will be discussed.
• New processor technology innovation helping to accelerate AI workloads
• Overview of competitors in AI workload acceleration
• Datacenter AI processor market forecast

Part 2: Revolutionizing the Market: The Unstoppable Rise of AI Accelerators - Julius Lo 

• The importance of TCO in achieving sustainable success for data center AI inference. The use of ASICs is highlighted as the most practical solution for specific purposes, including performance, power consumption, and energy efficiency.
• Introduction of RecAccel™ N3000, a purpose-built recommendation accelerator, which is designed to optimize TCO in data centers.
• The key features of RecAccel™ N3000, including its software-driven design, software stack, hardware DSA (Data Stream Accelerator), and IP design. 

Part 3: - Pervasive AI: Enabling Next-Generation Inference from the Data Center to the Edge - Sarina Sit 

As AI becomes ubiquitous to data center and enterprise computing and more companies integrate AI as a strategic tool to propel business forward and open new revenue streams, computing performance has become even more critical. In this session, Sarina will explore the myriad of applications for AI inference, share her view on the hardware and software portfolio that can address this space, and describe how AI inference accelerators can provide performance uplift for dedicated recommendation use cases.