Azure SQL Hyperscale tops Aurora PostgreSQL in Principled Technologies test

Principled Technologies, in testing commissioned by Microsoft, found Azure SQL Database Hyperscale outperformed Amazon Aurora PostgreSQL in most configurations it measured. The strongest results came at 16 vCores, where Azure SQL delivered up to 52% higher performance and 51% better performance per dollar.

Why it matters: - The results suggest Azure SQL Database Hyperscale can offer a stronger mix of speed and cost efficiency for organizations choosing between managed database services. - The comparison matters most for transactional workloads that need both high throughput and predictable cloud spend.

What happened: - Principled Technologies benchmarked Azure SQL Database Hyperscale against Amazon Aurora PostgreSQL in a third-party test commissioned by Microsoft. - The testing used the open-source HammerDB tool with the TPROC-C workload. - PT evaluated both services at 16 and 32 vCores across multiple concurrency loads. - The report measured raw transaction throughput and performance per dollar.

The details: - At 16 vCores, Azure SQL Database Hyperscale delivered up to 52% higher performance than Amazon Aurora PostgreSQL. - At 16 vCores, Azure SQL Database Hyperscale delivered up to 51% better performance per dollar. - Azure SQL Database Hyperscale outperformed Amazon Aurora PostgreSQL at nearly all concurrency loads tested at 16 vCores. - At 32 vCores, Azure SQL Database Hyperscale matched or exceeded Amazon Aurora PostgreSQL across all loads tested. - At 32 vCores, Azure SQL Database Hyperscale posted up to 12% better performance and performance per dollar. - The test used a 25,000-warehouse TPROC-C schema and an I/O-intensive workload designed to stress each service's storage subsystem. - PT ran each configuration three times and reported the median result. - The full report, Scale your most critical workloads with Azure SQL Database Hyperscale, includes detailed results and methodology. - Readers can also review the science behind the report for step-by-step configuration and testing details.

Between the lines: - The strongest showing came at the smaller 16-vCore configuration, which suggests Azure SQL Database Hyperscale's efficiency may be most noticeable before workloads scale up further. - Because the study was commissioned by Microsoft, the benchmark is useful as a product comparison but should be read with that relationship in mind. - The report frames database choice as a tradeoff between easy scaling, performance, and budget discipline.

What's next: - Organizations comparing managed database platforms may use the report's vCore and concurrency data as a starting point for their own workload testing. - Principled Technologies says the report includes enough detail for readers to replicate the configuration and evaluate the findings against their own needs.

The bottom line: - In this test, Azure SQL Database Hyperscale beat Amazon Aurora PostgreSQL on performance and cost efficiency in most scenarios, with its biggest advantage at 16 vCores.