Daisytuner Daisytuner
Coming soon
Continuous Benchmarking
Autotuning

Struggling with Slow Code? You're Not Alone.

Daisytuner is the platform to benchmark and optimize your code with precision. Leverage cutting-edge tools built by leading performance experts and compiler engineers.

GitHub Actions

Measure Everything, Right from Your CI.

Trigger benchmarking jobs across a wide range of processors. Our high-performance benchmarking cluster spans ARM, x86, GPUs, and specialized accelerators — ensuring your code runs optimally on any hardware.

CPU GPU Accelerators
Ampere Altra Max M128-30 NVIDIA L4 Tenstorrent Grayskull E75, Axelera AI Metis
AMD EPYC Genoa 9554 NVIDIA A16 Tenstorrent Grayskull E150, Axelera AI Metis
Intel Xeon Silver 4516Y+ NVIDIA A10 Tenstorrent Grayskull E150, Axelera AI Metis
Flame Graphs
Roofline Modeling
Cache Analysis

Unit Testing for Performance.

Define benchmarks for your code and catch performance regressions effortlessly with every commit. Track runtime changes or explore performance bounds from advanced performance models.

Daisytuner
VS Code LLVM

Want to Get Faster? Ask Our Co-Pilot.

Effortlessly apply optimizations like parallelization, tiling, and vectorization. Powered by advanced compiler analysis, our Co-Pilot delivers precise, high-performance code edits.

Coming soon
Daisytuner

Compiler + Machine Learning = ❤

Our Co-Pilot taps into the largest search engine for performance optimization.
Are you an expert in autotuning, GPU or FPGA kernels? Reach out to us and make it available for others.

Ready to Supercharge Your Code?

Join our waitlist and get exclusive early access to our platform — plus free benchmarking minutes on our high-performance cluster. Start optimizing, accelerating, and scaling your code effortlessly.

Get early access to the platform!