The Role of Advanced Computer Architectures in Accelerating Artificial Intelligence Workloads

2511.10010v1 cs.AR, cs.AI, cs.LG 2025-11-15
Авторы:

Shahid Amin, Syed Pervez Hussnain Shah

Abstract

The remarkable progress in Artificial Intelligence (AI) is foundation-ally linked to a concurrent revolution in computer architecture. As AI models, particularly Deep Neural Networks (DNNs), have grown in complexity, their massive computational demands have pushed traditional architectures to their limits. This paper provides a structured review of this co-evolution, analyzing the architectural landscape designed to accelerate modern AI workloads. We explore the dominant architectural paradigms Graphics Processing Units (GPUs), Appli-cation-Specific Integrated Circuits (ASICs), and Field-Programmable Gate Ar-rays (FPGAs) by breaking down their design philosophies, key features, and per-formance trade-offs. The core principles essential for performance and energy efficiency, including dataflow optimization, advanced memory hierarchies, spar-sity, and quantization, are analyzed. Furthermore, this paper looks ahead to emerging technologies such as Processing-in-Memory (PIM) and neuromorphic computing, which may redefine future computation. By synthesizing architec-tural principles with quantitative performance data from industry-standard benchmarks, this survey presents a comprehensive picture of the AI accelerator landscape. We conclude that AI and computer architecture are in a symbiotic relationship, where hardware-software co-design is no longer an optimization but a necessity for future progress in computing.

Ссылки и действия

Связанные статьи

eIQ Neutron: Redefining Edge-AI Inference with Integrated NPU and Compiler Innov...

#### Контекст Активное развитие интеллектуальных систем на основе глубоких нейронных сетей приводит к появлению новых т...

2025-09-20

LEGO: Spatial Accelerator Generation and Optimization for Tensor Applications

## Контекст Современные тензорные приложения, особенно фундаментальные модели и приложения в области генеративного ИИ, т...

2025-09-17

HPD: Hybrid Projection Decomposition for Robust State Space Models on Analog CIM...

## Контекст State Space Models (SSMs) являются эффективными альтернативами традиционным последовательностным моделям в о...

2025-08-19