Last week was Part 1 of our conversation with Akram’s Razor where he gave his account of how the crypto crash of 2018 impacted Nvidia and AMD. This week, Akram and James go through Nvidia’s data center business in depth and assess the growing competitive landscape for AI accelerators, from cloud and startup companies alike. Akram shares more about how he approached his research for his thesis on Nvidia and we get into a serious discussion about Google’s AI advantage and why he believes they are not making their TPU training chips freely available. We then debate Nvidia’s pricing for their data center boards and talk about their CEO, Jensen Huang, and his refreshing transparency and openness.