Paper Title: Bring Intelligence to Your Database Storage

Paper Abstract: Fast data that needs to be analyzed for business value often resides on Flash storage managed by a database. As businesses accumulate petabytes of data tracking their customer's every move, optimizing how database queries begins with how the data is stored, and what tools are used to quickly and accurately process the maximum amount of data in the minimum amount of time. Computational Storage has emerged as the leading solution to cohesively unlock both storage I/O and compute related bottlenecks, specifically in database functions that demand massive data movement and intense compute such as full table scans query requests. This presentation will review how modern Computational Storage easily integrates into database infrastructure at the application/API level. It will also present real-world benefits of reduced time-to-insight through specifically saving data movement for query processing by bringing this compute function directly to the storage drive.

Paper Author: Tong Zhang, Chief Scientist, ScaleFlux

Author Bio: Tong Zhang is co-founder and Chief Scientist at ScaleFlux, responsible for developing new approaches for computational storage products and exploring use cases in mainstream applications such as databases. He is also a professor at Rensselaer Polytechnic Institute, where his research areas include databases, filesystems, solid-state storage, digital signal processing, error correction coding, and device and computer architecture. He has published over 150 technical papers including many presented at prestigious USENIX/IEEE/ACM conferences or appearing in widely cited journals. He made pioneering contributions to the use of signal processing in flash memory systems and to the increased use of low-density parity-check (LDPC) codes in storage. He has over 20 patents either issued or pending. He earned a PhD in electrical and computer engineering from the University of Minnesota. He has been a presenter at several past Flash Memory Summits.