In this episode, Kevin Bowers returns to explain Jump's expansion from trading to core infrastructure, centered on what he calls the "great inversion": the real bottleneck in tech isn't compute, but data and I/O. He introduces Shelby, a new storage network, as a direct challenge to the "Hotel California for Data" model used by cloud providers. This same focus on efficient data flow—not just processing power—was the key to scaling Solana with Fire Dancer. Finally, Kevin explains how FPGAs from high-frequency trading are the critical hardware solution, allowing blockchains to bypass software's inefficient "Tower of Babel" and "get close to the wire" for true high performance. 00:00 - Expanding Beyond Trading and the Vision for Shelby 02:31 - Challenges in Storage and Data Management 04:37 - Building High-Performance Systems 08:04 - The Evolution of Jump's Technology 11:55 - The Economics of Cloud Storage 29:07 - Fire Dancer and Frankendancer 42:03 - The Cost of Optimization 42:48 - Machine Learning and Custom Networks 43:47 - Project Prioritization and Entropy 46:48 - Challenges in High-Performance Computing 56:38 - The Role of FPGAs in Trading and Blockchain 01:13:45 - Future of Hardware Acceleration in Blockchain 01:18:54 -Conclusion and Final Thoughts Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
