Capture the opportunity
A home’s rooftop solar and battery create an opportunity for compute after household energy needs are accounted for.
What if your rooftop could power the next big idea? We’re connecting solar homes with GPU computing to make more of Australia’s energy.
Four bedrooms, three bathrooms, a furnished balcony and a landscaped garden. The small GPU box lives inside the ground-floor utility room.
Solar energy and AI demand belong in the same conversation. SunStack brings them together, one home and one workload at a time.
A home’s rooftop solar and battery create an opportunity for compute after household energy needs are accounted for.
The SunStack platform matches suitable GPU jobs to available nodes, energy conditions and workload requirements.
Compute users get useful AI capacity. Homeowners can earn a share of the work hosted at their property.
Explore Australian GPU capacity for AI inference, experimentation and flexible workloads.
Explore GPU ComputeDiscover how hosting a GPU node could turn spare solar and battery energy into additional value.
Explore home hostingConnect the conditions of each home with the needs of each job. Make decisions across solar, storage, GPU performance and household priorities.
Meet the management platformRooftop energy changes through the day. Many AI tasks can be scheduled. Connecting the two creates a different way to think about infrastructure.
Solar generation and household demand do not always arrive together. Local computing introduces another possible use for energy that is available after the home’s needs are considered.
AI workloads have different memory, latency and continuity needs. SunStack’s direction is to match suitable jobs with available hardware and the energy conditions around it.
In this example, available generation serves household demand, a bounded GPU job and battery charging.
A simplified energy balance, not a forecast or a claim about hosting income. Change the hour to explore.
CONTRIBUTESA suitable place, connectivity and available energy.
RECEIVESVisibility, control and a proposed share of hosting value.
CONTRIBUTESA defined workload, data requirements and service needs.
RECEIVESUseful GPU capacity with an Australian deployment focus.
CONTRIBUTESCoordination, metering and the management layer.
RECEIVESA model for operating and developing a distributed compute network.
Commercial participation terms are agreed for each deployment. Research and staged trials inform the model.