The Future of Compilers
is Self-Learning
We envision a world where software automatically adapts to hardware, unlocking unprecedented performance and efficiency without human intervention.
A New Efficiency Layer for Computing
Static compilers are a thing of the past. A self-learning compiler analyzes workloads and hardware, tailoring optimizations to exact needs. The technology can be deployed on local developer machines to boost productivity, and scaled all the way to large data centers, transforming them into adaptive, self-optimizing compute fabrics.
Workload Specific
One-size-fits-all solutions are inefficient. Boost efficiency with individual optimization.
Full Portability
Seamlessly switch between different processors without relying on the vendor's support for your algorithms.
Latest Updates
Dr. Nora Hagmeyer lectures on compilers for novel processors at TU Darmstadt
Daisytuner joins the lecture series "The Future of Embedded Systems" at TU Darmstadt.
Daisytuner and ETH Zurich present at CGO 2025 in Las Vegas
Daisytuner and ETH Zurich present a novel approach to verifiable AI for compiler optimization at CGO 2025.
Daisytuner and SPCL win HiPEAC Technology Transfer Award in Barcelona
Daisytuner and SPCL receive the HiPEAC Technology Transfer Award for transferring cutting-edge compiler research to industry.
