Idaho National Laboratory - Sovereign AI Energy Campus
A planning-grade blueprint for a sovereign, energy-integrated AI campus conceptually situated on DOE-managed lands at Idaho National Laboratory.
The project frames INL's nuclear research heritage, integrated energy systems leadership, and regional transmission context as inputs for high-discipline proxy
modeling rather than operational assumptions. The architecture emphasizes phased growth, strict mission isolation, environmental caution, and reversible decision
gates, enabling DOE to explore how advanced AI workloads might be evaluated alongside future energy system evolution without committing to deployment, procurement,
or facility control. This site serves as the portfolio's anchor for long-horizon energy-AI co-planning under conservative governance.
View Project Document →
Oak Ridge - AI Supercomputing & Energy Convergence Campus
A strategic planning concept focused on the convergence of frontier AI, high-performance computing adjacency, and energy-aware campus design within the Oak Ridge
ecosystem. The project explores how AI workloads, scientific computing, and regional energy context can be evaluated together in proxy models to support national
research missions, advanced analytics, and infrastructure resilience planning. Emphasis is placed on modular growth patterns, disciplined governance, cybersecurity
separation, and grid-aware workload modeling, while remaining fully unclassified and non-operational. Oak Ridge functions as the portfolio's compute-intensive
innovation hub, balancing scientific ambition with infrastructure realism.
View Project Document →
Savannah River Site - Sovereign AI Enclave
A conservative, governance-first planning framework for introducing a small, tightly controlled sovereign AI enclave within the Savannah River Site's highly
sensitive nuclear and environmental mission environment. The concept treats the Site Use Permit process as a primary architectural governor, embedding environmental
protection, mission coexistence, and reversibility directly into the campus lifecycle. AI workloads are modeled strictly in proxy space using unclassified datasets
to support stewardship analytics, long-horizon planning, and systems modeling without interfacing with operational facilities. This site demonstrates how advanced AI
infrastructure could coexist with national security missions under maximum regulatory discipline.
View Project Document →
Paducah - Ultra-Scale Brownfield Power Nexus
A high-discipline planning construct examining the potential reuse of a large industrial brownfield footprint for very large AI infrastructure footprints in conceptual
models. The project leverages publicly documented historical design context and logistics corridors as proxy planning references only, without assuming current capacity
or readiness. The architecture focuses on modular siting, phased redevelopment, remediation coexistence, and infrastructure auditability, enabling DOE to explore how
legacy industrial assets might support future energy-intensive computing under strict environmental and governance constraints. Paducah provides the portfolio's testbed
for scale economics, brownfield reuse complexity, and long-horizon infrastructure transformation.
View Project Document →
Pacific Northwest National Laboratory - Grid-Edge AI Resilience Node
A planning-grade concept exploring how AI infrastructure could function as a grid-edge analytical and resilience node within a regional energy research environment.
The project models interactions between AI workloads, grid stress scenarios, storage abstractions, and cybersecurity separation using proxy datasets and unclassified
methods. Emphasis is placed on resilience analytics, outage modeling, and infrastructure protection research rather than operational control or deployment. PNNL serves
as the portfolio's systems-resilience laboratory, linking AI to grid reliability, digital infrastructure protection, and regional energy coordination planning.
View Project Document →
NREL Flatirons - AI Renewable Systems Nexus
A specialized planning environment focused on AI interaction with high-variability renewable energy systems, flexible compute scheduling, and advanced cooling and
efficiency modeling. The project uses proxy datasets to examine how AI workloads could adapt to renewable intermittency, storage constraints, and evolving grid
conditions without asserting operational integration or physical deployment. Flatirons functions as the portfolio's renewable experimentation platform, generating
transferable insights for modular data center design, energy-aware orchestration, and future sustainability strategies across the broader DOE landscape.
View Project Document →
