Steve Orrin on the Importance of Hardware in AI Development

On this episode of Data Driven, our hosts Andy Leonard, BAILeY, and Frank La Vigne are joined by guest Steven Orrin, an expert in software and hardware innovation at Intel.

The episode dives into the crucial role that hardware plays in AI development, from data curation to training and inferencing. Steve emphasizes the importance of hardware optimization for specific workloads to achieve powerful and timely training. They also explore the impact of hardware on inferencing, particularly in real-time applications like autonomous driving. Intel, as Steve explains, is providing a diverse set of hardware architectures, including CPUs, AI accelerators, and edge AI chips, to address various AI workloads.

In addition to discussing hardware, the conversation touches on several other interesting topics. Steve dives into the concept of collapsing data related to planes or vehicles into a single integer level or bit, highlighting the value of data but also cautioning against excessive data collection. They also touch on the future of edge computing, the challenges of achieving high performance and energy efficiency, the advantages of a diverse team, and Intel's commitment to collaboration and open source.

Overall, this episode provides valuable insights into the importance of hardware in AI development and offers a fascinating glimpse into Intel's approach to providing hardware solutions for different AI workloads. Join us on this episode of Data Driven to learn more about the role of hardware in AI and Intel's contributions to the field.