LLM-Driven Design Space Exploration of FPGA-based Accelerators

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Sharma, Vinamra, Fu, Xingjian, Haris, Jude, Cano, José
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917467057553408
author Sharma, Vinamra
Fu, Xingjian
Haris, Jude
Cano, José
author_facet Sharma, Vinamra
Fu, Xingjian
Haris, Jude
Cano, José
contents Designing field-programmable gate array (FPGA)-based accelerators for modern artificial intelligence workloads requires navigating a large and complex hardware design space encompassing architectural parameters, dataflow strategies, and memory hierarchies, making the process time-consuming and resource-intensive. While the SECDA methodology enables rapid hardware-software co-design of accelerators through SystemC simulation and FPGA execution, identifying optimal accelerator configurations still requires substantial manual effort and domain expertise. This work presents SECDA-DSE, a framework that integrates Large Language Models (LLMs) into the SECDA ecosystem, comprising tools built around SECDA to automate the design space exploration (DSE) of FPGA-based accelerators. SECDA-DSE combines a structured DSE Explorer for generating accelerator configurations with an LLM Stack that performs reasoning-guided exploration using retrieval-augmented generation and chain-of-thought prompting, alongside a feedback loop that enables reinforced fine-tuning for continuous improvement. We demonstrate the feasibility of SECDA-DSE through an initial high-level synthesis based evaluation of a generated accelerator design that meets synthesis timing and resource constraints on an Zynq-7000 FPGA.
format Preprint
id arxiv_https___arxiv_org_abs_2605_05920
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LLM-Driven Design Space Exploration of FPGA-based Accelerators
Sharma, Vinamra
Fu, Xingjian
Haris, Jude
Cano, José
Hardware Architecture
Artificial Intelligence
Performance
Designing field-programmable gate array (FPGA)-based accelerators for modern artificial intelligence workloads requires navigating a large and complex hardware design space encompassing architectural parameters, dataflow strategies, and memory hierarchies, making the process time-consuming and resource-intensive. While the SECDA methodology enables rapid hardware-software co-design of accelerators through SystemC simulation and FPGA execution, identifying optimal accelerator configurations still requires substantial manual effort and domain expertise. This work presents SECDA-DSE, a framework that integrates Large Language Models (LLMs) into the SECDA ecosystem, comprising tools built around SECDA to automate the design space exploration (DSE) of FPGA-based accelerators. SECDA-DSE combines a structured DSE Explorer for generating accelerator configurations with an LLM Stack that performs reasoning-guided exploration using retrieval-augmented generation and chain-of-thought prompting, alongside a feedback loop that enables reinforced fine-tuning for continuous improvement. We demonstrate the feasibility of SECDA-DSE through an initial high-level synthesis based evaluation of a generated accelerator design that meets synthesis timing and resource constraints on an Zynq-7000 FPGA.
title LLM-Driven Design Space Exploration of FPGA-based Accelerators
topic Hardware Architecture
Artificial Intelligence
Performance
url https://arxiv.org/abs/2605.05920