AI

TACS-RF: Tool-driven Agentic Circuit Synthesis of RF and mmWave Modules

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 26/10/2026, 04:03:24 UTC.

Briefings

A new framework called TACS-RF uses an LLM agent to automate the design of RF and mmWave modules, achieving a 240-251 GHz sub-THz transmitter in a 90 nm SiGe BiCMOS process. The agent autonomously designed critical components like a 60 GHz push-push VCO and a differential bowtie antenna, delivering +5.1 to +6.3 dBm output power and phase noise of -80 to -88 dBc/Hz at 1 MHz offset while consuming only 361 mW DC power.

Key points

• The TACS-RF framework employs a large-language-model (LLM) coding agent named Claude Opus 4.8 to directly drive industry-standard EDA tools including Cadence Virtuoso, Spectre, Calibre, Cadence EMX, and Ansys HFSS.
• The framework successfully designed a 240-251 GHz sub-THz transmitter in a 90 nm SiGe BiCMOS process, with the agent semi-autonomously creating a 60 GHz push-push Colpitts VCO, multistage differential cascode power drivers, and a differential bowtie antenna.

Context

This research represents a significant step toward a 'text-to-ASIC' future for designing RF/mmWave hardware by automating the synthesis of complex sub-THz modules using an LLM-driven agent.

Limitations

Validation was limited to a 90 nm SiGe BiCMOS process; the framework's performance and scalability beyond this specific process node remain unverified.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 26/10/2026, 06:48:08 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeihszvku6nz4muzh56rwufxmqgshqf5xxk35csckgbvpz2hcm64gsi