Phase 01 · Evaluate
Hybrid numerical + AI evaluation
Four physics engines returning machine-readable margins over real power grids and 2.5D/3D stacks, callable thousands of times inside your flow.
NoveetyAI was founded to turn three decades of published VLSI reliability, thermal and computation research into production chiplet design software — independently implemented, engineered for speed, and callable by an autonomous agent.
Advanced packaging changed the shape of the problem. In a 2.5D or 3D chiplet system-on-package, electrical, thermal, reliability and mechanical physics are coupled — IR-drop drives Joule heating, heat drives electromigration and thermomigration, thermal mismatch drives warpage. Analyzed separately and checked last, each answer is misleading.
We build the simulation engines that are accurate enough to sign off on and fast enough to sit inside an optimization loop — then we make them callable, so an autonomous design agent can evaluate thousands of candidates against real physics rather than a surrogate.
Every engine ships with a ground-truth reference solver, and every result is validated against it.
Order-of-magnitude acceleration is what makes closed-loop optimization possible at all.
CLI-first, structured I/O, headless by default. No GUI in the loop.
Support for OpenROAD alongside commercial flows keeps the door open for research and small teams.
The industry has converged on this. Writing for Semiconductor Digest, Synopsys' Larry Williams argues that stacked architectures, chiplets and 3DIC packaging have made electrical, thermal and mechanical behavior impossible to separate — effects once treated as second-order now govern performance, yield and reliability, and physics has to enter at design time rather than wait for downstream sign-off.
"The physics is coupled."
That is the premise NoveetyAI is built on: in an advanced package, a flow that analyzes each domain in isolation is describing a chip that doesn't exist.
Larry Williams · Synopsys "Silicon Meets Reality: Why Physics Is Now a First-Order Design Constraint for Chips" — Semiconductor Digest ↗And the cost of getting it wrong is not theoretical. Industry analysis attributed NVIDIA's Blackwell delay to a thermal-mechanical failure in TSMC's CoWoS-L package: coefficient-of-thermal-expansion mismatch between the GPU chiplets, LSI bridges, RDL interposer and substrate. The reported remedy was a redesign of the top metal layers and bump structures, plus requalification before mass production could resume.
"…causes warpage and failure of the whole SiP."
Warpage is not a late cosmetic check. On the most closely watched silicon in the industry, it moved a schedule — and it is precisely the physics WarpStack computes.
Phil Garrou · Packaging IFTLE 607 "Why Nvidia's Blackwell is Having Issues with TSMC CoWoS-L Technology" — 3D InCites, October 2024 ↗Recognizing the problem is not the same as solving it. Moving physics to design time only works if the physics is fast enough to be called on every candidate — and if something is driving those calls. Sign-off simulators are accurate but far too slow to sit in an optimization loop, so integrity still gets checked last, when it is most expensive to fix. NoveetyAI closes both halves: multiphysics engines built for loop speed, and an agentic flow that acts on what they return.
No matter how fast the flow becomes, or how much of it an agent drives, hardware is ultimately limited and gated by fundamental physics. Signal integrity, power and thermal integrity, electromigration and thermal-mechanical aging are not soft objectives an optimizer can trade away — they decide whether a part works at all, and whether it still works years later in the field.
That is the role NoveetyAI's sign-off capability plays in an agentic EDA flow. An agent may explore thousands of candidates with fast surrogates, but every candidate it advances is checked against physics-grade engines validated on a ground-truth reference solver. Speed moves the search; physics decides what ships.
Prof. Tan directs the VLSI System and Computation Lab (VSCLAB) at UC Riverside, where the NoveetyAI engines originated. His research spans VLSI reliability modeling, thermal analysis and management, hardware acceleration, and AI/LLM-driven approaches to electronic design automation.
NoveetyAI is advised by three strategic advisors from production EDA and open-source design automation — see the advisory board.
The algorithm behind each NoveetyAI engine was first published as peer-reviewed research from VSCLAB — open literature, available to anyone — and validated against a reference solver on industrial-scale benchmarks: IBM power-grid circuits, SAED32 place-and-routed designs, native IEEE 3Dblox chiplet packages, and a 2nm CoWoS assembly. We then built our own production implementation of each method: engineered for loop speed, CLI-first, structured I/O, and supported as a commercial product.
Geometry-resolved thermal solver for 2.5D/3D stacks on IEEE 3Dblox geometry, checked against a bundled 3D FEM reference.
MOR-model + 3D layer-wise FEM warpage. Validated on a 2nm CoWoS package, 200+ mechanical layers.
Advanced Krylov subspace reduction for transient power-grid analysis. Validated to 1.04M nodes.
The methods behind these engines were developed in academic research at the University of California, Riverside and published in the peer-reviewed literature, where they are available to the public. NoveetyAI's software is an independent implementation of those published methods — it is not licensed from, distributed by, or derived from software belonging to UCR. NoveetyAI is the trade name of Noveety AI, Inc., an independent company; UCR does not endorse its products or services.
The engines already make integrity and reliability observable at loop speed. Next, the agent acts on what it sees.
Four physics engines returning machine-readable margins over real power grids and 2.5D/3D stacks, callable thousands of times inside your flow.
The agent stops scoring candidates and starts proposing them — floorplan, power delivery, chiplet partition and stack-up optimized against real physics.
Design, optimization and sign-off in one agentic loop — from intent to a verified, integrity-clean hand-off.
The NoveetyAI suite is in active development. Tell us about your design and we'll get you set up — usually within a couple of business days.
Fill this in and we'll respond by email.