Related papers: SSR: A Stall Scheme Reducing Bubbles in Load-Use H…

The occurrence of bubbles in pipeline parallelism is an inherent limitation that can account for more than 40% of the large language model (LLM) training time and is one of the main reasons for the underutilization of GPU resources in LLM…

Modern general-purpose accelerators integrate a large number of programmable area- and energy-efficient processing elements (PEs), to deliver high performance while meeting stringent power delivery and thermal dissipation constraints. In…

WebRISC-V is a web-based educational tool designed to simulate the pipelined execution of assembly programs according to the RV64IM specifications (64-bit RISC-V processor). The tool allows users to investigate pipeline stalls, understand…

As inference workloads for large language models (LLMs) scale to meet growing user demand, pipeline parallelism (PP) has become a widely adopted strategy for multi-GPU deployment, particularly in cross-node setups, to improve key-value (KV)…

Our goal in this paper is to understand how to maximize energy efficiency when designing standard-ISA processor cores for subthreshold operation. We hence develop a custom subthreshold library and use it to synthesize the open-source RISC-V…

This work evaluates how well hardware-based approaches detect stack buffer overflow (SBO) attacks in RISC-V systems. We conducted simulations on the PULP platform and examined micro-architecture events using semi-supervised anomaly…

Pipeline parallelism is one of the key components for large-scale distributed training, yet its efficiency suffers from pipeline bubbles which were deemed inevitable. In this work, we introduce a scheduling strategy that, to our knowledge,…

RISC-V is a promising open-source architecture primarily targeted for embedded systems. Programs compiled using the RISC-V toolchain can run bare-metal on the system, and, as such, can be vulnerable to several memory corruption…

As RISC-V architectures proliferate across embedded and high-performance domains, developers face persistent challenges in performance optimization due to fragmented tooling, immature hardware features, and platform-specific defects. This…

It is intended in this document to introduce a handy systematic method for enumerating all possible data dependency cases that could occur between any two instructions that might happen to be processed at the same time at different stages…

On embedded processors that are increasingly equipped with multiple CPU cores, static hardware partitioning is an established means of consolidating and isolating workloads onto single chips. This architectural pattern is suitable for…

In computer architecture courses, we usually teach RISC processors using a five-stage pipeline, neglecting alternative organizations. This design choice, rooted in the 1980s technology, may not be optimal today, and it is certainly not the…

This paper presents a comprehensive analysis of the RISC-V instruction set architecture, focusing on its modular design, implementation challenges, and performance characteristics. We examine the RV32I base instruction set with extensions…

Load instructions often limit instruction-level parallelism (ILP) in modern processors due to data and resource dependences they cause. Prior techniques like Load Value Prediction (LVP) and Memory Renaming (MRN) mitigate load data…

Input buffered switches with Virtual Output Queueing (VOQ) can be unstable when presented with unbalanced loads. Existing scheduling algorithms, including iSLIP for Input Queued (IQ) switches and Round Robin (RR) for Combined Input and…

Pipeline parallelism has been widely explored, but most existing schedules lack a systematic methodology. In this paper, we propose a framework to decompose pipeline schedules as repeating a building block, and show that the lifespan of the…

The Instruction Set Architecture (ISA) defines processor operations and serves as the interface between hardware and software. As an open ISA, RISC-V lowers the barriers to processor design and encourages widespread adoption, but also…

Pipeline parallelism is widely used to scale the training of transformer-based large language models, various works have been done to improve its throughput and memory footprint. In this paper, we address a frequently overlooked issue: the…

Pipeline parallelism has emerged as a predominant approach for deploying large language models (LLMs) across distributed nodes, owing to its lower communication overhead compared to tensor parallelism. While demonstrating high throughput in…

It is customary to assess the reliability of underground oil and gas pipelines in the presence of excessive loading and corrosion effects to ensure a leak-free transport of hazardous materials. The main idea behind this reliability analysis…