The field of high-performance computing is moving towards increased heterogeneity, with a growing emphasis on specialized hardware and software co-design. Recent research has focused on optimizing performance, power consumption, and dependability in various applications, including artificial intelligence, machine learning, and data-intensive computing. Notable advancements include the development of new architectures, such as the Aurora supercomputer, and the exploration of novel computing paradigms, like neuromorphic computing and spiking neural networks. Additionally, there is a growing interest in leveraging emerging technologies, including SmartNICs, DPUs, and RISC-V-based systems, to accelerate data processing and improve overall system efficiency. Some noteworthy papers in this area include the proposal of THAPI, a tracing framework for heterogeneous APIs, and the development of SpikeStream, an optimization technique for spiking neural networks on RISC-V clusters. Overall, the field is experiencing a significant shift towards more specialized, efficient, and adaptable computing systems, driven by the increasing demands of modern applications and the need for sustainable, high-performance computing solutions.
Advancements in High-Performance Computing and AI
Sources
Performance Analysis of HPC applications on the Aurora Supercomputer: Exploring the Impact of HBM-Enabled Intel Xeon Max CPUs
Achieving Dependability of AI Execution with Radiation Hardened Processors
THAPI: Tracing Heterogeneous APIs
HiAER-Spike: Hardware-Software Co-Design for Large-Scale Reconfigurable Event-Driven Neuromorphic Computing
What Every Computer Scientist Needs To Know About Parallelization
MemPool Flavors: Between Versatility and Specialization in a RISC-V Manycore Cluster
A Survey on Heterogeneous Computing Using SmartNICs and Emerging Data Processing Units (Expanded Preprint)
Fused-Tiled Layers: Minimizing Data Movement on RISC-V SoCs with Software-Managed Caches
Work-In-Progress: Accelerating Numpy With OpenBLAS For Open-Source RISC-V Chips
RealProbe: An Automated and Lightweight Performance Profiler for In-FPGA Execution of High-Level Synthesis Designs
Exploring energy consumption of AI frameworks on a 64-core RV64 Server CPU
WebRISC-V: A 64-bit RISC-V Pipeline Simulator for Computer Architecture Classes
Learning Cache Coherence Traffic for NoC Routing Design
oneDAL Optimization for ARM Scalable Vector Extension: Maximizing Efficiency for High-Performance Data Science
Virtual memory for real-time systems using hPMP
FERIVer: An FPGA-assisted Emulated Framework for RTL Verification of RISC-V Processors
dpBento: Benchmarking DPUs for Data Processing
CVA6-VMRT: A Modular Approach Towards Time-Predictable Virtual Memory in a 64-bit Application Class RISC-V Processor
SpikeStream: Accelerating Spiking Neural Network Inference on RISC-V Clusters with Sparse Computation Extensions
Membrane: Accelerating Database Analytics with Bank-Level DRAM-PIM Filtering
Beyond Moore's Law: Harnessing the Redshift of Generative AI with Effective Hardware-Software Co-Design
Efficient Deployment of Spiking Neural Networks on SpiNNaker2 for DVS Gesture Recognition Using Neuromorphic Intermediate Representation