ARTEMIS X

We choose the hard things.

Rice University · September 12, 1962

ARTEMIS X
HomeCloudAI factoriesEnergyCompanyContact

750 MW
Planned program capacity, up to · 2030
July 2027
35 MW
Campus One live — 175 racks, 12,600 GPUs, in NVIDIA DSX POD units on infrastructure already standing
In place
~50 MW
Campus One expansion path within the same hall — no new construction
Q4 2027
50 MW
Campus Two online — a second identified campus on the same design
By 2030
Up to 750
The program — multiple campuses, one repeatable clean-sheet design

Planned capacity on the NVIDIA Vera Rubin NVL72 platform — 200 kW liquid-cooled racks. Figures are planned program capacity; projects are in development.

01 / The build

Go deeper

The cloud

Frontier compute, dedicated by design

Vera Rubin NVL72 at rack scale, bare metal on the NVIDIA reference fabric, factory-integrated NVIDIA DSX POD units, closed-loop cooling with zero water.

Explore the cloud ↗
AI factories

Megawatts, with dates

Campus One live July 2027, Campus Two 50 MW online Q4 2027, and the behind-the-meter power model they all share.

See the campuses ↗
Energy

Power first, by design

Off the grid queue, behind the meter, fueled by sea — with SCR on every engine.

See the energy model ↗
The company

We choose the hard things

The program to 750 MW, a team from hyperscale data centers, power development and infrastructure finance, and how to reach them.

Meet Artemis X ↗

Contact

Come build with us

Artemis X develops dedicated AI infrastructure. If you need capacity or want to build alongside us, reach the team directly — tell us the scale and the timeline. Builders: we are hiring — same address.

Contact Artemis X↗