Daisytuner provides highly integrated tools to measure, analyze, and track performance metrics of software applications. Our bot delivers status checks and automated reports for pull requests, ensuring you never miss a performance bug.

Catch Performance Bugs Before They Catch You

1. Trigger benchmarking jobs from workflow definitions in your repositories. Jobs run on dedicated processors inside lightweight containers—no virtual CPUs, no jitter from the infrastructure.
2. Measure everything to identify bottlenecks: from context switches to DRAM bandwidth, cache misses, and FLOP counters.
3. Struggling to interpret runtime data for call graphs? Our compiler extensions automatically instrument source code regions, providing performance insights directly for the lines in your code.

How Fast Is Your Code?

Runtime is important, but performance is key. Unlock state-of-the-art analyses and models to understand your code's performance:

• Scalability: Leverage compiler-driven analysis of parallelism alongside scalability measurements.
• Peak Performance: Compare memory bandwidth and FLOP measurements for your loops against processor peak capabilities.
• Energy Efficiency: Run your code on a variety of processors and evaluate its energy consumption across different architectures.

Getting Started

Requirements:

• Plan: The free plan includes a monthly quota of benchmarking minutes. For more minutes, upgrade your plan via our website.
• Availability: Waitlist.

Setup Process:

1. Install the app.
2. Add your first workflow to your repo. Checkout the examples for more info.
3. Make changes to your code and measure the performance.