Recap: Presto Summit SF 2019
July 1, 2019
What’s Presto Summit? It’s the leading Presto conference co-organized by our partner Starburst Data and the Presto Software Foundation.
Overview of the Summit
- Presto is among the fastest growing open source analytical query frameworks with production use cases across industries such as retail, telco, tech and more
- This was a full house event at Twitter HQ with more than 150 attendees
- Excellent keynote on the future of Presto delivered by Martin Traverso, Dain Sundstrom, and David Phillips – the co-creators of Presto and co-founders of the Presto Software Foundation (slides)
What We Learned
- Presto is high performance and has rich functionalities designed for interactive SQL queries
- Rise of cloud deployments of Presto both all cloud as well as hybrid cloud are growing. There is a need to simplify hybrid deployments which currently include making copies of data in the cloud and in some cases to HDFS and managing the Hadoop cluster.
- Big data workloads are increasingly looking for simpler compute orchestration and adopting Kubernetes. Starburst Presto announced a new Kubernetes operator to simplify deployment and scaling of clusters.
Reasons to try the Presto, Alluxio, and Any Storage Stack
- High query performance without operational overhead of data copying or ETL
- Query on data anywhere: hybrid, public cloud, on-premise
- Consistent and low latency
Learn more: Starburst Presto and Alluxio announce strategic OEM partnership | Presto with Alluxio | Download Alluxio
Additional resources:
- Community office hour (virtual): Building Fast SQL Analytics with Presto, Alluxio, and S3
- Got questions? Chat with Alluxio experts on Slack
.png)
Blog

MLPerf Storage v3.0: Alluxio's Distributed Cache Delivers 147 GiB/s Llama 3 Checkpoint Writes
Alluxio’s MLPerf Storage v3.0 results demonstrate that organizations can keep persistent AI data in Amazon S3 while delivering high-performance training and checkpointing through a distributed cache close to compute. Alluxio achieved 147.03 GiB/s checkpoint write bandwidth for Llama 3 405B and near-linear scaling from one to 32 workers.

