Quantum OS Kernel Research

VonX is building EigenOS — the first quantum-native operating system kernel. It orchestrates hybrid quantum-classical computation across simulators and real QPUs with deterministic scheduling, memory-safe execution, and real-time control.

Explore a mini, simple and working prototype at haulvisor.vercel.app.

EigenOS treats quantum processes with the same rigor as classical OSes treat processes, threads, and memory. It fills the essential system-level gap required to unlock scalable, production-grade quantum AI.

Focus on Quantum–Classical AI Models

Our work extends far beyond infrastructure. We’re developing models that leverage EigenOS to run hybrid architectures efficiently:

QuantumGPTMini, a token-to-qubit transformer using entanglement-based attention and variational feedforward networks.

Integration of EigenOS primitives enables low-latency hybrid inference, streaming responses from CPU→QPU→CPU.

These models prove EigenOS’s worth and enable rich experiments in hybrid AI.

A recent study by Amire Ramazan—“QuantumGPTMini: A Hybrid Quantum-Classical Transformer for Enhanced NLP”—documents the architecture and benchmark performance. Published April 28, 2025, it outlines how QuantumGPTMini uses about one million classical parameters alongside lightweight quantum circuits. Read the paper.