DOE's 12 Genesis Mission Phase II awards: $159 million, from a fusion twin to quantum magnets

The Energy Department put names and numbers on the Genesis Mission on October 8, 2026. It announced 12 new Phase II awards totaling $159 million and six Phase I awards, raising Phase II projects to 14 and the first-year cohort to 297. The department's Under Secretary for Science, Darío Gil, described the portfolio as "297 teams across all 50 states."

The projects are the useful part. The branding around them, "Super Intelligence for science," is the part that needs reading carefully.

The 12 projects

The department lists the new Phase II projects, all from its Office of Science. In its own descriptions:

• Commonwealth Fusion Systems: an AI-enabled digital twin of the SPARC fusion device, to simulate and optimize its operations.
• UC San Diego: a roughly fivefold larger RNA structure database to train models for engineering resilient crops and microbes.
• MIT: AI combined with supercomputing calculations of fundamental particle physics, known as Lattice QCD.
• UC Irvine: MAESTRO, an AI platform for mapping and managing deep underground reservoirs for geothermal energy.
• University of Illinois Urbana-Champaign: AXIS, which uses AI to design electrochemical processes for extracting rare-earth elements.
• University of Washington: AI tools to predict and design functional enzymes.
• Fermilab: AXESS, using AI to design microchips for extreme environments such as space and high radiation.
• Lawrence Berkeley National Laboratory: MOAT-Core, a shared platform for autonomously running and designing particle accelerators across the national labs.
• Harvard: ASQC, combining AI with quantum hardware to address quantum error correction.
• Argonne National Laboratory: AI4HPC, a framework for modernizing, optimizing and verifying supercomputing software.
• Oak Ridge National Laboratory: a physics-informed framework for designing quantum magnets from desired properties.
• Northeastern University: an automated assistant for operating and tuning the particle separator at the Facility for Rare Isotope Beams.

Two earlier Phase II projects round out the 14: GridFM 2.0 from the Office of Electricity, funded at $11.5 million, and Prometheus from the Office of Nuclear Energy.

What Phase II means

The program runs in two tiers. Phase I awards support proof-of-concept work, and Phase II turns the concepts into actual projects. That is a conventional research funding ladder. It is useful to know that 12 projects at $159 million averages roughly $13 million each, but the department publishes no per-project figure, so treat that division as arithmetic, not a fact about any single award.

The scale looks modest next to the capital flows covered elsewhere on this site. It is a rounding error against the spending described in the State of AI Report, which puts four hyperscalers at roughly $733 billion of 2026 capital spending. What Genesis buys is access to national laboratory facilities and a directed set of problems, not a rival to commercial compute.

The other announcements that day

Per Nextgov, the Energy Department announced the awards at the White House's AI Science Summit alongside several other items. A Genesis Mission graduate fellowship pilot carries up to $100 million in total funding and promises two hands-on research experiences, one at a national laboratory or user facility and one with industry. The 2026 Early Career Research Program selected 67 scientists for a combined $96 million, drawn from 39 universities, 11 national laboratories and user facilities, and 26 states.

The department also released eight priority scientific problems for quantum computing, building on its Quantum Genesis initiative launched in June, which aims to deploy a fault-tolerant, scientifically relevant quantum computer by 2028. A separate Quantum Genesis competition from September offers up to $215 million in planned funding for companies to deploy machines with at least 100 logical qubits.

Industry arrived at the same event. Nvidia pledged $1 billion over five years, and Anthropic pledged $150 million over three years. We covered the first in Nvidia's $1 billion commitment and the second in our separate report on Anthropic's pledge.

The label

The department's release is titled "Energy Department Announces New Genesis Mission Awards to Advance Super Intelligence for Science." Gil said the awards represent "a critical step in turning the promise of Super Intelligence for science." The phrase follows the administration's September 29 order, which we examined in our explainer on the Super Intelligence executive order.

Read the project list, though, and the technology is familiar: digital twins, surrogate models, machine-learning design tools, automation of instruments. Nothing in the release describes a system that is more capable than what labs already sell. The name signals a policy direction. It is not evidence of a new capability.

What is still unknown

• Per-project budgets. Only the aggregate is published.
• Results. The Genesis Mission launched in November 2025 and announced first projects in July. No finding is cited.
• Funding durability. The department's own chief of staff said in April that the out years are what it needs to solidify, according to FedScoop. The proposed fiscal 2027 budget carries $1.2 billion for AI supercomputers under a new Office of Artificial Intelligence and Quantum, while the Office of Science would fall 13 percent, if Congress approves it.

Our take

This is a competent, specific funding announcement with an inflated label on top. The projects, from the SPARC twin to the accelerator platform, are the kind of AI-for-science work that plausibly pays off. The word to watch is "Phase II": it means money, not results. We would judge Genesis on published findings by this time next year, and on whether the out-year funding the department says it needs actually arrives.