Compute that stays close.
Exploring useful AI workloads that can run on local hardware.
THE NEXT FRONTIER OF COMPUTE
From data centers to devices.
A new frontier for private, local-first AI.
01 / A DISTRIBUTED FUTURE
Across a data center. Inside a machine. On your desk. Intelligence is becoming part of the world around us—and where it runs matters.
Compute near a person, place or machine. The edge creates room for local inference, responsive experiences and greater control over where data travels.
PROXIMITY / RESPONSE / CONTROLData centers provide shared infrastructure for training, large models and connected services. They remain part of the picture as more intelligence moves outward.
CAPACITY / TRAINING / SHARED INFRASTRUCTURECameras, robots, personal computers and embedded systems bring inference into the physical world. Memory, power and software shape what each device can do.
CONTEXT / SENSING / LOCAL INFERENCEOne connected landscape. Each workload belongs where it works best.
02 / INTRODUCING ONCORE
Edge AI for data privacy & compute autonomy. OnCore is exploring a more personal relationship with the machines that power intelligence.
IN DEVELOPMENT
Local inference. Private storage. Hardware and software considered together, from the first layer.
An evolving engineering project for compact, local-first AI systems.
Follow the buildExploring useful AI workloads that can run on local hardware.
Making deliberate choices about storage, execution and what leaves the system.
Bringing compute, interfaces and compact hardware into the same design process.
03 / THE TECHNOLOGY BEHIND IT
Explore compact computers, accelerators and embedded platforms—with their trade-offs in view.
SOURCE-LINKED SPECIFICATIONS ↗Local model references and their requirements. A starting point for understanding what can run where.
FEATURED LOCAL AI MODELS ↗04 / FOLLOW YOUR CURIOSITY
Understand the technology shaping personal AI. Tell us which direction matters to you.
THE SIGNAL CONTINUES
Hardware. Ideas. Engineering in the open.
Follow OnCore updatesCheck out the YouTube channel: @EdgePlease