polito.it
Politecnico di Torino (logo)

Accelerating Real-Time Edge AI: Unraveling the potential of the VE2302 in the AMD landscape

Lorenzo Radaele

Accelerating Real-Time Edge AI: Unraveling the potential of the VE2302 in the AMD landscape.

Rel. Bartolomeo Montrucchio. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Informatica (Computer Engineering), 2024

[img]
Preview
PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (6MB) | Preview
Abstract:

In the fast-evolving landscape of Artificial Intelligence (AI) and machine learning, real-time applications demand cutting-edge solutions. Xilinx, now under AMD, aims to address this need with the Versal™ AI Edge Series. This series promises high-performance, low-latency AI inference for diverse applications, including automated driving, factory automation, and healthcare systems. The Vek280, a potent board from this series, showcases impressive performance but faces challenges due to its high production cost. This thesis delves into the development and potential impact of the VE2302, a more affordable and accessible alternative designed to strike a balance between superior performance and cost-effectiveness. The VE2302 targets industries where real-time AI applications play a crucial role, presenting opportunities for innovation and broader market reach. The exploration includes a comprehensive review of the Versal Family, focusing on the Versal AI Edge Series, and detailed insights into the VE2302's design, interfaces, and potential applications. Benchmark evaluations against existing boards and a comparative analysis underscore the VE2302's capabilities, highlighting its significance in the evolving landscape of edge AI devices. The strategic shift toward affordable, high-performance solutions positions Xilinx/AMD competitively in the burgeoning real-time application domain, opening avenues for diverse and impactful applications.

Relators: Bartolomeo Montrucchio
Academic year: 2023/24
Publication type: Electronic
Number of Pages: 82
Subjects:
Corso di laurea: Corso di laurea magistrale in Ingegneria Informatica (Computer Engineering)
Classe di laurea: New organization > Master science > LM-32 - COMPUTER SYSTEMS ENGINEERING
Aziende collaboratrici: AVNET EMG ITALY SRL
URI: http://webthesis.biblio.polito.it/id/eprint/31002
Modify record (reserved for operators) Modify record (reserved for operators)