Ikioma Mini
The desk-sized entry point — two linked nodes, one private stack.
The deskside supercomputer — up to 1T params in one memory pool.
Hardware, delivered and deployed
Your data never leaves the floor. Air-gappable, no egress, no telemetry — the trust boundary has a serial number you can point at in an audit.
Own the machine and agents run around the clock. No per-token meter, no overage invoices — you pay for electricity, not tokens.
Hardware is a capital asset, not a subscription line. Weights, tunes, and stack stay yours — perpetually, with trade-in credit on next-gen silicon.
A superchip with coherent memory — GPU and CPU sharing one pool, so a trillion-parameter model runs without sharding or offload. Models you used to rent in the cloud, running on a machine under your desk.
Partition the machine into isolated instances, each with its own memory and compute — your data scientists, engineers, and agents run side by side without stepping on each other. Nobody shares their work.
Built for always-on, self-evolving agents — long-running, autonomous, and private to the building. No token meter, no rate limits; and when one box isn't enough, link two at 800 Gb/s.
Out-of-band management, hardware root of trust, secure boot — the visibility your IT team expects, in a fleet of one or a fleet of many. A dedicated engineering team, a 4-hour SLA, and updates that are signed and reversible.
Assembled in Yokohama and Eindhoven. Boards burn in for 96 hours before they leave the floor. Every system carries a serial, a calibration sheet, and a name etched on the chassis — because the people who built it stand behind it for a decade, not until next quarter's earnings call.
Preinstalled and pre-validated: the ikioma inference runtime, your choice of optimized open models, and the production middle layer (audit log, RBAC, evals) configured out of the box.
Loading 3D view…
The deskside superchip carries coherent memory, twin memory stacks and a dedicated cooling block. Drag to orbit, or explode the assembly to see how the machine is laid out.
Illustrative 3D approximation — not manufacturing CAD. Schematic dimensions only.
Mini for the desk, Ikioma for the team, Pro for the fleet. Every version ships the same model and inference stack — the scale is what changes.