DOE GENESIS MISSION · SEMICONDUCTORS & MICROELECTRONICS

VERA

Verification Enhancement through Reasoning Agents

A four-agent AI framework that automates semiconductor design verification —
from specification parsing to coverage closure to hardware security validation.

60–70% of chip dev cycles
consumed by DV today
$100M cost of a single
tape-out re-spin
50% target reduction
in DV cycle time
$294M DOE Genesis Mission
total funding pool
Explore the Framework ↓ Contact the Team

Design verification is
America's chip bottleneck.

Modern processors contain tens of billions of transistors. Proving they work correctly before manufacturing consumes the majority of engineering time — and current tools haven't fundamentally changed in two decades.

WHERE CHIP DEVELOPMENT TIME GOES

Design 15% Verification 65% ← VERA targets this Physical 12% Validation 8% $10M–$100M per tape-out error
01

Coverage Closure Is Manual and Slow

Reaching 95%+ functional coverage on a complex IP block takes 3–8 weeks of iterative constraint tuning. Engineers re-discover the same coverage gaps design cycle after design cycle.

02

Testbench Creation Is Hand-Coded

UVM testbench components — sequences, scoreboards, coverage groups — are largely written from scratch. A single complex IP verification effort can require 100,000+ lines of testbench code.

03

Failure Triage Consumes Expert Time

When a simulation fails, isolating the root cause takes 4–40 hours of a senior engineer's time per failure. At scale, this bottleneck alone can stall a tapeout schedule by weeks.

04

Security Validation Has No Automation

Hardware Trojans, privilege escalation, and side-channel vulnerabilities are manually checked — if at all. No systematic AI tooling exists for RTL-level hardware security verification.

Four agents. One
unified knowledge graph.

VERA is a modular AI platform where four specialized agents share a persistent Verification Knowledge Graph (VKG), enabling institutional learning across design cycles for the first time.

RTL Source SystemVerilog / VHDL Spec Documents NL / PDF / Word SIA Specification Intelligence Agent LLM · Fine-tuned RAG · Embedding Search TSA Testbench Synthesis Agent Code Gen LLM · AST Self-Correction Loop CCA Coverage Closure Agent Deep RL · GNN Coverage Prediction FTA Failure Triage Agent LLM Triage · LSTM Bug DB Retrieval VERIFICATION KNOWLEDGE GRAPH VKG Design Intent Cov. State Sim History Bug Records Sec. Props. Simulation Env. Verilator · VCS · Xcelium REST API Integration DVE · vManager · Questa Data flow to VKG VKG knowledge node
🔍
AGENT 01 · SIA

Specification Intelligence Agent

Ingests natural-language specs, RTL source, and existing plans. Produces structured, machine-readable verification intent with full requirement traceability, including implicit security constraints.

Fine-tuned LLM RAG NLP Security Extraction Embedding Search
⚙️
AGENT 02 · TSA

Testbench Synthesis Agent

Generates syntactically correct, executable UVM testbench components and SVA security property checkers from verification intent. Iterative self-correction loop handles compiler errors automatically.

Code Gen LLM AST Validation Self-Correction SVA Generation
🎯
AGENT 03 · CCA

Coverage Closure Agent

The core RL subsystem. Continuously analyzes coverage database state during simulation and adaptively generates test stimuli to close functional and security coverage gaps. Replaces manual constraint tuning.

Deep RL GNN on RTL Coverage Prediction
🔬
AGENT 04 · FTA

Failure Triage Agent

Activates on simulation failures. Classifies bug type, distinguishes functional bugs from hardware security violations, and proposes root-cause corrections. Hardware Trojan indicator detection included.

LLM Triage LSTM Waveform Bug DB Retrieval
VERIFICATION KNOWLEDGE GRAPH (VKG)

All four agents share a persistent, versioned knowledge graph spanning design intent, coverage state, simulation history, security properties, and bug records. The VKG enables institutional learning that current DV environments cannot provide — knowledge compounds across design cycles.

Measurable outcomes,
not research promises.

Every VERA target is benchmarked against current industry practice on the CVA6 open-source RISC-V processor — a production-quality design representative of commercial IP verification challenges.

COVERAGE CLOSURE: TRADITIONAL vs. VERA

100% 75% 50% 25% 95% target Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Traditional (3–8 weeks) VERA (target: 50% faster)

Automation Coverage Targets

UVM Testbench Components Auto-Generated 70%
Security Assertions Auto-Generated 60%
Spec-to-Plan Accuracy (SIA) 80%
Bug Classification Accuracy (FTA) 85%

Time and Compute Reduction Targets

Coverage Closure Cycle Reduction 50%
Verification Plan Creation Time Reduction 80%
Failure Triage Time Reduction (functional) 75%
Testbench Dev Time Reduction 40%

Industry expertise meets
AI engineering execution.

🔧
Douglas Lane, MBA, PMP
VeAssis LLC · Managing Director
35+ years semiconductor DV at Intel (DFx team lead), AMD, and IBM. Builder of ScopeBase.AI, a live multi-agent AI platform on Microsoft Commercial Marketplace. Leads AI platform development and program management for VERA.
DBEWBEMBEHUBACDBE
🛡️
Michael Taborn
Independent · Security Verification Architect
Intel systems architect focused on security and validation. Post-Intel work at NIH on health and life sciences computing infrastructure. UT Austin alumnus. Leads VERA's hardware security verification capability and security assertion library design.
Intel Alum NIH Research
🏗️
Christopher Brown
Mavros Technology LLC · Lead Technical Architect
Enterprise systems integration and cloud-native infrastructure architect. Leads VERA's EDA tool integration layer, REST API architecture, and Verification Knowledge Graph deployment infrastructure.
DBE NCTRCA

Built for four Genesis
Mission challenges.

GENESIS CHALLENGE VERA CONTRIBUTION
AI-Driven Autonomous Laboratories VERA automates the full DV workflow end-to-end, enabling research teams to execute orders of magnitude more simulation experiments with minimal manual intervention — the definition of an autonomous scientific laboratory for chip design.
Semiconductors & Microelectronics Primary focus: VERA directly addresses the dominant productivity bottleneck in U.S. semiconductor design programs, with an explicit hardware security verification extension that no existing AI tool provides.
National Security (NNSA) Hardware security validation capabilities address NNSA requirements for verified, Trojan-free chip designs used in national security systems. The security assertion library will be released as an open DOE platform resource.
Advanced Manufacturing Faster, more thorough DV reduces tape-out failures and re-spins, directly improving yield and competitiveness in domestic semiconductor manufacturing programs supported by CHIPS Act investment.

Nine months.
Four parallel workstreams.

WORKSTREAM
MONTHS 1–5
MONTHS 6–9
WS1
Environment
& Data
MILESTONE 1CVA6 simulation environment operational. HDL, UVM, and security assertion corpora assembled and annotated.
MILESTONE 2Simulation Trace Dataset v1.0 released as open-access DOE platform resource.
WS2
SIA &
TSA
MILESTONE 3SIA: 80% spec-to-plan accuracy. Security NLP pipeline operational on CWE/CAPEC hardware entries.
MILESTONE 4TSA: 70% UVM component / 60% security assertion generation. Peer-reviewed paper submitted.
WS3
CCA
MILESTONE 5RL environment stable. Baseline DRL policy trained on 10 CVA6 scenario configurations.
MILESTONE 6CCA achieves 40%+ simulation cycle reduction on CVA6 coverage closure benchmark.
WS4
FTA &
Integration
MILESTONE 7FTA: 85%+ bug classification accuracy. Security violation classification pipeline operational.
MILESTONE 8Full VERA integration on CVA6. VKG live. Phase II letter of intent prepared.

Ready to advance
U.S. semiconductor leadership?

VERA is seeking a university partner with active semiconductor or EDA research infrastructure.
Phase I awards range from $500K–$750K over 9 months (DE-FOA-0003612).

Contact Doug Lane · doug@veassis.com
VeAssis LLC · 110 N IH 35, Ste 315 #3295, Round Rock, TX 78681
UEI: M3V5YJJ62QL5 · EIN: 93-1970104
Certifications: DBE · WBE · MBE · HUB · ACDBE (40+ jurisdictions)