Intel / article

Can Agents Design Better Chips with a Higher Level Abstraction?

Source: arxiv.org
Author: Zijian Ding, Yang Zou, Yizhou Sun, Jason Cong
Date: Published Sep 21, 2026

Context

AI chip-design agents can benefit from a compiler doing part of the design work. In the AHRR study, an first writes C++ describing a computation. High-level synthesis converts it into a circuit description, which the agent then refines at the lower level. This combines compiler design knowledge with opportunities for detailed optimization.

The researchers report a 2.62× geometric-mean improvement in calculated hardware execution time across 13 eligible task-and-model comparisons against direct circuit-description generation. Execution time combines simulated cycles with timing after circuit placement and routing. Each experiment ran once, so the result supports this workflow on the tested programmable-hardware tasks without establishing a universal advantage.

Key takeaways · AI-distilled

The study compares four workflows - Direct RTL Design, Agent-based HLS Design, Post-Compiler HLS Refinement, and Post-HLS RTL Refinement - then combines agent-based HLS with RTL refinement into what it calls AHRR.
AHRR's 2.6x geometric-mean speedup over direct RTL design was measured across an 11-task suite, using FPGAs as the platform.
Case studies attribute the gain to two complementary effects: HLS distills design knowledge into abstractions agents can actually use, while the later RTL-refinement pass recovers lower-level optimizations HLS alone misses.
The authors say the design-flow tradeoffs they studied are largely independent of the target technology, even though FPGAs, chosen for ease of end-to-end deployment, were the only platform tested.

Terms in this piece · Glossary

AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.

Read the source arxiv.org