Abstract: In various industries, in-database analytics are crucial for decision-making. Yet, the growing amount of data presents challenges as traditional methods become excessively time-consuming and costly. Moore's Law and Dennard scaling, which previously aided data analytics workloads, are reaching their physical limits, and a new approach is needed. In this talk, I'll present a new perspective on the data representation substrate for data platforms. I argue that a bit-level data shredding approach could offer significant advantages for future data platforms. In many respects, the argument here is that our transition from a row-store to a column-store didn’t go far enough. Another crucial step is required, and we need to shed further to create bit-stores. Further, for higher performance, this approach has the benefit of being amenable to implementation in hardware with low (area, delay, and power) costs, and I’ll present early results pointing in this direction.

Bio: Jignesh Patel is a professor in the Computer Science Department at Carnegie Mellon University. His research focuses on data management, emphasizing both system efficiency (e.g., scalable data platforms) and human efficiency (e.g., designing LLM-based query interfaces). His papers have been recognized as the best papers at top database conferences, including SIGMOD and VLDB. He is a fellow of the AAAS, ACM, and IEEE organizations. He has also received teaching awards at the U. of Wisconsin and the U. of Michigan, and he is the co-founder of four startups.

Website: https://jigneshpatel.org/

Twitter/X: @pateljm