Dynamic Data Structures on the GPU

Hello CUDA community,

We're honoured to host John Owens at our next online meetup!
On March 17th (19:00 UTC+3), he'll talk about "Dynamic Data Structures on the GPU". We will add a link to the Zoom meeting closer to the event so stay tuned.

Speaker:
John Owens

Abstract:
Dynamic data structures allow updates to the data structure without having to rebuild it completely. I'll discuss five dynamic GPU data structures that we designed and built—log-structured merge trees, quotient filters, linked lists and hash tables built atop them, dynamic graphs, and B-trees. The talk is less about the specifics of these data structures and more about principles that we followed in building them and what we learned from the experience.

Agenda:

• Welcome message
• Demonstration of CLion for CUDA development
• Raffle of free CLion licenses
• The talk
• Post meetup discussions

About the speaker:
Dr. Owens is interested in building programmable and high-performance parallel computer systems. His work primarily concentrates on the "graphics processing unit" (GPU). His research group focuses on parallel primitives applications across a wide variety of computing domains, including graph analytics, parallel data structure and algorithm design and implementation, computer graphics, scalable computing, automotive computing, numerical computing, machine learning, and programming languages. This work aims to design, implement, and analyze the building blocks of future parallel computing systems.