The History and Future of Servers and the Cloud with Oxide's Bryan Cantrill

The history of servers, the cloud, and what’s next – with Oxide

TLDR: Bryan Cantrill, CTO of Oxide Computer, joins The Pragmatic Engineer podcast to discuss the evolution of servers and the cloud over the past 30 years. He shares insights from his time at Sun Microsystems during the dot-com era, the rise of AWS, and his current work at Oxide building modern server infrastructure.

Summary: In this episode of The Pragmatic Engineer podcast, Bryan Cantrill offers a fascinating historical perspective on the evolution of computing infrastructure. He argues that the dot-com bust, contrary to popular belief, was a period of greater technical innovation than the boom that preceded it. With fewer resources, engineering teams were forced to be more creative and focused, leading to breakthroughs like ZFS and DTrace at Sun Microsystems.

Cantrill also provides an inside look at the early days of AWS and how Jeff Bezos's pricing strategy was a "genius move" that masked the profitability of the cloud business. He explains how AWS S3 was effectively subsidizing Amazon's retail operations, and how this made it difficult for competitors like Joyent to compete.

The conversation then turns to the rise of Kubernetes and its role in providing "cloud neutrality." Cantrill argues that Kubernetes gained traction in part because it offered a way for companies to avoid being locked into the AWS ecosystem.

The episode also delves into Cantrill's work at Oxide Computer, where he and his team are building a new type of computer from scratch. He discusses the challenges of hiring hardware engineers who are willing to work from a "clean sheet of paper" rather than relying on reference designs.

For anyone interested in the history of computing, the economics of the cloud, or the future of server infrastructure, this podcast is a must-listen. It provides a rare, long-range view that connects the past, present, and future of our industry.

Key takeaways:

• The dot-com bust was a period of intense innovation in systems software.
• AWS's early pricing strategy was a brilliant business move that disguised the profitability of the cloud.
• Kubernetes emerged as a way to achieve "cloud neutrality" and avoid vendor lock-in.
• Building hardware from first principles requires a different kind of engineering mindset.

Link: The history of servers, the cloud, and what’s next – with Oxide