Advancements in High-Performance Computing and Numerical Analysis

The field of high-performance computing and numerical analysis is advancing rapidly, with a focus on developing innovative solutions to complex problems. Recent developments have centered around improving the efficiency and accuracy of numerical simulations, particularly in the context of large-scale systems and complex applications. One notable trend is the increasing use of specialized architectures and compilation strategies to accelerate computationally intensive tasks, such as combinational logic computation and sequential logic synchronization. Additionally, there is a growing interest in developing novel approaches for computing floating-point errors and optimizing system-of-systems modeling and optimization. The development of new tools and methodologies for system-level design, such as ORCHA, is also enabling more efficient and effective exploration of possible configurations for complex systems. Noteworthy papers include: CCSS, which achieves significant speedup over state-of-the-art multi-core simulators through specialized architecture and compilation strategies. PI-detector, which efficiently computes floating-point errors by injecting small perturbations into the operands of individual atomic operations. ORCHA, which enables the exploration of possible configurations for complex systems through a unified interface layer and code generation and synthesis.

Sources

CCSS: Hardware-Accelerated RTL Simulation with Fast Combinational Logic Computing and Sequential Logic Synchronization

Computing Floating-Point Errors by Injecting Perturbations

System-of-systems Modeling and Optimization: An Integrated Framework for Intermodal Mobility

ORCHA -- A Performance Portability System for Post-Exascale Systems

Tools and Methodologies for System-Level Design

Universal Synthesis of Differentiably Tunable Numerical Abstract Transformers

Dual-Numbers Reverse AD for Functional Array Languages

Performance Portable Gradient Computations Using Source Transformation

Built with on top of