NeuReality, a developer of AI infrastructure solutions, has introduced NR-NEXUS, an inference operating system designed to power large-scale AI deployments. Already deployed with beta customers, the platform aims to transform fragmented systems into production-ready “AI token factories,” enabling organizations to optimize inference workloads across diverse hardware environments.
Built on NeuReality’s expertise in AI hardware architecture and large-scale inference system design, NR-NEXUS is a hardware-agnostic operating system that supports CPUs, GPUs, network interface cards (NICs), and emerging XPUs. The platform is designed to unify the AI inference stack, addressing inefficiencies in underutilized GPUs and siloed infrastructure that often increase costs and reduce performance.
By orchestrating inference workloads across hyperscale cloud environments, dedicated GPU clusters, and other emerging hardware, NR-NEXUS stabilizes performance, improves utilization, and lowers the cost per token generated. The system allows organizations to scale operations without re-architecting existing deployments, enabling a seamless transition to more efficient, enterprise-scale AI factories.
“AI inference is rapidly becoming one of the largest computing markets in the world, yet the infrastructure stack around it remains fragmented,” said Moshe Tanach, CEO of NeuReality. “With NR-NEXUS, we are defining the operating system for AI token factories – enabling organizations to run and scale inference workloads efficiently across GPUs, emerging XPUs, hyperscalers, and dedicated AI clusters. As open-source models and AI-native applications proliferate, operators need infrastructure that gives them flexibility rather than lock-in. NR-NEXUS provides that foundation.”
The platform is aimed at NeoCloud providers, enterprises, and semiconductor vendors seeking to consolidate fragmented infrastructure into unified inference platforms. NeuReality positions NR-NEXUS as a tool to accelerate time to market for new AI models while maximizing the return on investment from AI factory builds.
Founded in 2019, NeuReality develops purpose-built inference infrastructure, including NR2 AI-SuperNIC, NR1 AI-CPU, and NR1 Inference Appliance. The company employs 80 people across facilities in Israel, Poland, and the U.S. NR-NEXUS follows an open, standards-based approach, making it compatible with a broad range of hardware.
