Benchmark Go performance with Sweet and Benchstat
Introduction
Overview
Launch an Arm-based c4a-standard-4 instance
Launch an Intel Emerald Rapids c4-standard-8 instance
Install Go, Sweet, and Benchstat
Benchmark types and metrics
Manually run benchmarks
Manually run Benchstat
Install the automated benchmark and Benchstat runner
Run the automated benchmark and Benchstat runner
Next Steps
Benchmark Go performance with Sweet and Benchstat
Introduction
Overview
Launch an Arm-based c4a-standard-4 instance
Launch an Intel Emerald Rapids c4-standard-8 instance
Install Go, Sweet, and Benchstat
Benchmark types and metrics
Manually run benchmarks
Manually run Benchstat
Install the automated benchmark and Benchstat runner
Run the automated benchmark and Benchstat runner
Next Steps
Review frequently asked questions
These FAQs were drafted with an approved AI-assisted workflow and reviewed by Arm contributors before publication. Human technical review remains part of the process so the final page reflects engineering rigor, accuracy, and Arm editorial standards.
On the VM instances page, confirm that the series and machine type is
c4a-standard-4 for the Arm-based VM and c4-standard-8 for the Intel-based VM.Yes. The comparison relies on running the same Go Benchmarks natively on each architecture and then using Benchstat to analyze the two result sets.
Run the installer from your home directory, which creates a
$HOME/benchmarks/sweet path. Verify that the directory exists before you proceed to run benchmarks.Run Sweet on each VM to execute the benchmarks and generate raw performance data. After both runs complete, use Benchstat to compare the results across the two systems.
Inspect
sec/op for execution time, average-RSS-bytes for average resident memory, peak-RSS-bytes for maximum resident memory, and peak-VM-bytes for maximum virtual memory. Lower values indicate less time or memory for these metrics.