The GI/ITG International Conference on Architecture of Computing Systems (ARCS) conferences series has over 38 years of tradition reporting leading edge research in computer architectures and operating systems. This year, the conference has a special focus on: "Open-Source Architectures: From RISC-V to AI Accelerators"

New chips are designed either as general-purpose devices with a wide user portfolio in mind, or target the needs of specific applications (e.g. machine learning pipelines for deep learning workloads, mobile or embedded systems). Accordingly, they might rely on well-established architectural approaches (CPU-like, vector extensions), or bring completely new disruptive ideas limiting energy consumption. In this landscape, ARCS stands out in hosting and presenting such concepts and ideas for versatile architectures. Nowadays, such architectures often include open-source solutions such as RISC-V or AI accelerators, which are drawing more and more attention in academia and industry. ARCS’26 aims to leverage this growing interest in such architectures to foster their development, performance, and dissemination. We welcome all contributions on hardware architectures, as well as their programming models, software stacks (operating systems, compilers...), their insertion into computing systems, and the challenges to port optimized code.

In addition to the main conference topics, ARCS 2026 will host:

• A Special Track on Organic Computing.
• A Special Track on Dependability in the Era of Hardware-Accelerated AI Systems
• A PhD Forum to present the work of young doctoral students (10 page papers).
• A Group Forum for young professors, researchers, and consortia to present their research (5 page papers).

• Hardware Architectures
  • Reconfigurable architectures
  • System-on-chip
  • Parallel and distributed systems
  • High-performance systems
  • Heterogeneous many-core architectures
  • Domain-specific accelerator architectures
  • Real-time and mixed-criticality systems
  • I/O and non-volatile memory
  • Energy-efficient architectures
  • Smart network technologies
  • RISC-V and open-source architectures
  • AI accelerators
  • Dependable and fault-tolerant architectures

• Programming Models and Runtime Environments
  • Programming models for heterogeneous computing
  • Tools for power monitoring and performance analysis
  • Operating systems, algorithms, and data structures for heterogeneous architectures
  • Hypervisors and middleware for homogeneous and heterogeneous multi-/many-core computing platforms
  • System management including but not limited to scheduling, memory management, power/thermal management, and RTOS
  • Domain-specific languages and programming models
  • Architecture-specific code generation and optimization
  • Architectural simulation and emulation

• Cross-sectional Topics
  • Near- and in-memory computing
  • Memory and network compression
  • Organic Computing
  • Pervasive systems
  • Autonomous systems
  • Approximate Computing
  • Support for safety and security
  • Hardware in the loop
  • Optimization and performance analysis