Manycore system‑on‑chip (MCSoC) architectures have become central to modern computing—from embedded and mobile platforms to autonomous systems, edge‑AI accelerators, and high‑performance computing. Although hardware capabilities continue to scale in parallelism, heterogeneity, and energy efficiency, the software stack capable of fully exploiting these architectures still lags behind. Delivering predictable performance, real‑time behavior, and efficient resource utilization on deeply parallel SoCs remains a demanding challenge for both industry and academia.
As MCSoC platforms integrate diverse processing elements, complex memory hierarchies, and sophisticated on‑chip communication fabrics, traditional programming approaches are no longer sufficient. Developers must adopt new programming abstractions, compilation strategies, runtime systems, and optimization techniques that can bridge the gap between hardware potential and practical software performance. This track invites contributions that advance programming methodologies, tools, and frameworks specifically designed for embedded multicore and manycore systems.
Topics of interest include, but are not limited to:
- Compilers and compiler optimizations for MCSoCs
- Automatic code generation techniques
- Cross‑assemblers and toolchain support
- Programming models for heterogeneous and parallel SoCs
- Memory management and data‑movement optimization
- Runtime systems and resource management
- Object‑oriented approaches for embedded parallelism
- Concurrent and parallel software design
- Secure programming techniques and cryptography for MCSoCs
Formers Chairs
- Prof. Stéphane Louise, CEA, LIST, France (14th IEEE MCSoC-2021, 15th IEEE MCSoC-2022, 16th IEEE MCSoC-2023)
