Roofline is a German RWTH Aachen spin-off building compiler and runtime software for deploying AI models on edge devices. Its SDK translates models from frameworks such as PyTorch, TensorFlow and ONNX into executables optimised for different CPUs, GPUs and neural processors. The company develops a retargetable compiler based on MLIR and IREE, a system-on-chip inference runtime and performance tools so product and chip developers can deploy language, vision, audio and other models across heterogeneous hardware.
