Speaker:

Benjamin Owad

Date:

Tue Nov 18, 2025 @ 12:00pm EST

Date:

Tue Nov 18, 2025

Time:

12:00pm EST

Location:

GHC 8115

Title:

Optimizing the Table Scan Operator: I/O Minimization and Runtime Adaptivity

System:

Snowflake

---

Talk Info:

Table scan is a foundational operator in any analytical database and is often the primary bottleneck for a given query. This talk provides a technical deep dive into optimizations our team has developed for the table scan operator. First, we will discuss I/O reduction techniques, including pruning strategies to avoid reading unnecessary data and storage request coalescing to batch I/O operations efficiently. Next, we will explore runtime adaptivity, with features like adaptive scan, a methodology to balance CPU and memory usage with I/O bandwidth, dynamic column reordering, and data encoding schemes.

Bio:

Benjamin Owad is a computer science enthusiast based in the San Francisco Bay Area working at Snowflake. He studied computer science at Carnegie Mellon University, concentrating in systems. His interests span databases, open-source software, vintage hardware, and niche operating systems