S1E4 - Databricks, Snowflake and the return of benchmarking

 

In this episode, Terry, Tori & Craig discuss why it is not common to see benchmarking in database systems. In November 2021, Databricks and Snowflake went head-to-head when Databricks published their winning TPC-DS benchmark, which included a benchmark against Snowflake, which sparked a back-and-forth. From the outside it looked like a squabble between vendors, but what it started, has implications far beyond Databricks & Snowflake. Tune in to the great Database Benchmark War. We also talk about David Dewitt, the problems with Image Inference without context and the 2022 ACM Turing award winner.

Links:

Databricks Sets Official Data Warehousing Performance Record - The Databricks Blog

Eliminating the Anti-competitive DeWitt Clause for Database Benchmarking - The Databricks Blog

Industry Benchmarks and Competing with Integrity - Snowflake Blog

Snowflake Claims Similar Price/Performance to Databricks, but Not So Fast! - The Databricks Blog

TPC-DS Homepage

Eliminating the DeWitt Clause for Greater Transparency in Benchmarking (singlestore.com)

 

Listen to full episode :

Terry McCann

Microsoft MVP. Principal Consultant and Owner of Advancing Analytics Limited, an Advanced Analytics consultancy in the UK. Advancing Analytics helps businesses advance their analytical capabilities. Our focus is on Data Science, Data Engineering, DataOps and applied AI. Terry holds a Master's degree in Data Science - with a focus on DataOps for Machine Learning. Organiser of the Data Science Exeter user group, frequent speaker at conferences across the world and the host of the Data Science in Production Podcast.

https://www.advancinganalytics.co.uk
Previous
Previous

S1E5 - The one where Terry, Tori & Craig explore the Data Lakehouse

Next
Next

S1E3 - Getting back to conferences