Back to feed
Dev.to
Dev.to
7/16/2026
Benchmarking Postgres vs ClickHouse for 100M LLM Trace Rows on a Mac Mini

Benchmarking Postgres vs ClickHouse for 100M LLM Trace Rows on a Mac Mini

Original: Can a Mac Mini Handle 100 Million Rows?

Short summary

A developer benchmarked Postgres vs ClickHouse on a Mac Mini M4 Pro using up to 100 million rows of simulated LLM trace data, mirroring Langfuse's migration story. Both databases handled the full dataset without crashing, but performance diverged significantly on analytics queries. The experiment provides practical guidance for teams deciding whether to move LLM observability workloads from Postgres to a columnar store.

  • •Both Postgres and ClickHouse handled 100M rows of LLM traces on a Mac Mini M4 Pro without crashing
  • •ClickHouse acquired Langfuse in early 2026, motivating the real-world comparison
  • •Performance gaps emerge on dashboard aggregations and metrics queries, not raw data capacity

Generated with AI, which can make mistakes.

#ai-tools#industry-adoption#open-source
Read full article at Dev.to

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

The original title is "Postgres vs ClickHouse? I use both. 4 tricks from the split."
Dev.to
Postgres or ClickHouse? Row vs Column Storage, and When Each Wins
Dev.to
Can ClickHouse Replace a Vector Database?
Dev.to
Hetzner was cheaper at every size I tested and I still chose managed Postgres
Dev.to