CODAI Workshop Summary
The CODAI Workshop at HiPEAC 2025 is a dynamic gathering of experts from the AI compiler and AI accelerator communities, bridging academic and industrial research to explore the latest innovations in AI. The workshop invites original research and industry applications on a range of topics, including:
- Compilers for AI: Topics like partitioning, mapping, retargeting, intermediate representations, and domain-specific languages for heterogeneous systems (e.g., μC, DSP).
- Compiler Optimization for Performance: Techniques such as exploiting memory hierarchy, offloading, and asynchronous execution.
- Compiler Optimization for Efficiency and Speed: Methods like compression, pruning, quantization, neural architecture search, and auto-tuning.
- Performance Estimation for Compilers: Tools like virtual prototyping, surrogate modeling, and profiling.
- Code Generation & Hardware Backends: Focus on embedded AI accelerators, particularly RISC-V-based and beyond-von-Neumann platforms.
- Application Case Studies: Real-world examples of compilers for vision, sensor processing, brain-inspired algorithms for Edge AI, and on-device learning.
This workshop is an essential event for those advancing AI compilers and systems.
For more infos visit: https://www.codai-workshop.com