Why We Put GPUs in the Database
Presented by:
Montana Low
I grew up on a farm in Oklahoma, wired a lumberyard to host my first server, and moved to the West Coast 20 years ago. I’m lucky to enjoy solving complex problems. I start from first principles to make sure I can finish with quality. I’ve been privileged to know people who’ve taught me far more than I could have learned on my own. Now I'm building a company of people who also pursue interesting things in life.
The intersection of hardware advancements and machine learning (ML) innovations has deep implications for databases. The integration of GPUs into Postgres is a response to the growing demand for real-time, sophisticated data processing, enabling advanced analytics, machine learning model training, and inference with increasing tight latency constraints against increasingly large scale data. In this talk, we’ll discuss next gen hardware and data advancements, including the challenges faced, lessons learned, and the future outlook.
- Date:
- 2024 April 17 09:40 PDT
- Duration:
- 20 min
- Room:
- Almaden
- Conference:
- Postgres Conference 2024
- Language:
- English
- Track:
- Ops
- Difficulty:
- Intermediate