As architectures grow in complexity and power/performance requirements evolve far beyond what traditional computing facilities offer, advances in compiler technology are paramount to enable the construction of performant, power-efficient, and resilient application payloads. In this era of expanding architectural possibilities, this workshop seeks new research in compilers tailored for traditional architectures but also for the unique challenges posed by physical computing systems that perform computations based on their analog behavior. In addition to compiler technologies enabling new features (co-processors, accelerated instructions, tensor memory architecture engines, etc.) in traditional architectures (CPUs, GPUs, CGRAs, FPGAs), this year, in CFW, we seek to explore new compilation techniques for the intricacies inherent in emerging computing paradigms, such as analog computing, neuromorphic computing, near and in-memory processing, chemical reaction networks, among others, expanding the discussion to creative solutions when compiling for physical computing systems. This workshop serves as a gathering ground for researchers and experts from both industry and academia, fostering the presentation and exploration of the latest advances in Compilers for Computing Frontiers.

Topics of interest include, but are not limited to:

• Compiler extensions for HPC, Data Analytics, IoT, Cloud, and Edge computing
• Compiler design for low-power environments
• Compiler design for extreme parallelism
• Compiler optimizations and optimization frameworks
• Compiler optimizations for high-level abstractions
• Extensions to support analog computing
• Extensions to support neuromorphic computing
• Extensions to support computation with chemical reaction networks (CRNs)
• Extensions for emerging memory technologies (near and in-memory computing)
• Extensions to existing compiler frameworks: CLANG, GCC, MLIR, LLVM, etc.
• Template meta-programming constructs
• Domain-specific language compilers
• Compiler re-targeting for non-traditional/special-purpose architectures
• Compiler extensions for AI/ML architectures
• Application of ML to compiler technologies
• Compiler design for hardware description languages
• Compiler extensions for High-Level Synthesis tools targeting FPGAs or ASICs
• Compiler infrastructure to support new features in traditional architectures (CPUs, GPUs, CGRAs, FPGAs, etc.)

Submission Guidelines

This year, we are accepting two types of submissions: Full Papers and Extended Abstracts.

Full Papers: Authors are encouraged to submit full papers of up to 6 pages, excluding references, describing novel work.

Extended Abstracts: Authors are welcome to submit short papers describing early-stage, in-progress, and/or exploratory work to elicit feedback, discover collaboration opportunities, and spark discussion. Extended abstracts should be up to 2 pages, including references. Please, during submission, prefix your title with the string “EA:” to indicate that you are submitting an extended abstract.

Important: All accepted papers will be published in the workshop section of the Computing Frontiers proceedings.

All paper submissions must adhere to the official ACM conference format. As the review process is double-blind, removal of all identifying information from paper submissions is required (i.e., cite own work in third person). Papers not conforming to the above submission policies on formatting, page limits, and the removal of identifying information will be automatically rejected. Authors are strongly advised to submit their papers with the final list of authors, as changes may not be feasible later.

All papers must be submitted through the official submission website: Easychair

Committees

Technical Program Committee

• Serena Curzel - Politecnico di Milano
• Jose Monsalve Diaz - Argonne National Laboratory
• Joseph Manzano - Pacific Northwest National Laboratory
• Nicolas Bohm Agostini - Pacific Northwest National Laboratory, Northeastern University
• Antonino Tumeo - Pacific Northwest National Laboratory

Steering Committee

• David Donofrio - Tactical Computing Laboratories
• John Leidel - Tactical Computing Laboratories

Contact

• Nicolas Bohm Agostini – nicolas [dot] agostini [at] pnnl [dot] gov