Rising chip demand and costly tape-outs push new entrants; winners pair physics-grounded methods with fab/test data loops and veteran EDA talent.








Get your own research: Email to agent@olymposhq.com with subject "fetch" and link in body
















Recent advances in RTL-focused EDA assistants are being driven by fully open-source, non‑commercial methods that already show measurable improvements over earlier LLM baselines, signaling weak differentiation for proprietary offerings in this space arxiv.org.
Even with integrated error-detection and correction, reported end-to-end accuracy tops out at 91% on representative RTL tasks—insufficient for a safety- and correctness-critical workflow that still demands human-in-the-loop review and rigorous integration with established verification flows arxiv.org ewadirect.com.
Waveform-conditioned agents that propose targeted RTL edits based on simulation traces also risk misreading complex waveforms, producing plausible but incorrect changes that create hard-to-detect functional bugs without stringent verification ewadirect.com arxiv.org arxiv.org.
Get your own research: Email to agent@olymposhq.com with subject "fetch" and link in body.
(Pre-seed to Series A)
Learn more about Olympos and how our VC services can transform your sourcing process
Have more questions about Olympos?
Contact Us