Google has appointed Amin Vahdat as the chief technologist for AI infrastructure, a newly established role reporting directly to CEO Sundar Pichai. This move highlights the growing importance of AI development within the company, as Google is projected to invest up to $93 billion in capital expenditures by the end of 2025, with the figure expected to rise substantially next year.
Vahdat, a seasoned computer scientist with a PhD from UC Berkeley, has been instrumental in shaping Google’s AI foundation over the past 15 years. Prior to joining Google in 2010, he held academic positions at Duke University and UC San Diego, with an impressive research background and numerous published papers focused on enhancing computer efficiency at scale.
At a recent Google Cloud event, Vahdat introduced the cutting-edge seventh-generation TPU named Ironwood, highlighting its remarkable specifications: with over 9,000 chips per pod delivering 42.5 exaflops of compute power, surpassing the capabilities of the world’s leading supercomputer at that time. Vahdat emphasized the exponential growth in demand for AI compute, indicating a surge of 100 million times over just eight years.
Behind the scenes, Vahdat has been spearheading crucial work that bolsters Google’s competitiveness, including the development of custom TPU chips for AI operations and the implementation of the high-speed Jupiter network facilitating seamless data transfer across Google’s servers.
Source: TechCrunch
Leave a Reply