Behind the scenes of LENZO’s development is a hands-on, iteration-driven engineering environment.
At NAIST’s Computing Architecture Group in Nara, our team is building, testing, and scaling LENZO Core — step by step, board by board.
LENZO Core is developed here through continuous iteration, moving from early board-level prototypes to increasingly complex, multi-board systems operating under real compute load.
As the architecture evolves, each new configuration introduces higher core density, greater system complexity, and sustained utilization. All testing is done on live hardware, allowing performance, power behavior, and stability to be evaluated directly rather than through simulation alone.
One key milestone shown during this phase is a live demonstration of ChatGPT running on LENZO Core. This end-to-end workload highlights the flexibility of the architecture and its ability to support modern AI applications even at an early development stage.
Every setup shown here represents a step forward — grounding LENZO Core’s design in real-world constraints and building toward a scalable, general-purpose compute platform.