Next-generation research centre advancing the frontiers of neural network architecture, cognitive systems, and synthetic intelligence.
At CNNRC, our researchers pursue fundamental breakthroughs in how machines learn, reason, and interact with the world. We span three interconnected research pillars.
Designing novel neural network topologies that push beyond the transformer paradigm. Our research explores spiking neural networks, liquid state machines, and neuromorphic computing models that more closely mirror biological cognition.
Active Models
Parameters Trained
Developing architectures for machine reasoning, memory-augmented networks, and hierarchical temporal memory. We focus on building systems that can form abstract concepts, transfer knowledge across domains, and exhibit genuine understanding.
Published Papers
Active Collaborations
Researching the path from narrow AI to general cognitive capabilities. Our work spans multi-modal learning, embodied cognition, self-supervised representation learning, and the theoretical foundations of machine consciousness.
Research Fellows
Active Projects
Every breakthrough at CNNRC is powered by infrastructure designed from the ground up for large-scale neural network research. Our facilities represent one of the densest concentrations of AI-optimised compute in the world.
Twelve thousand NVIDIA H100 nodes interconnected via NVLink and InfiniBand, delivering 85 exaflops of mixed-precision compute for training models at unprecedented scale.
Custom-designed optical interconnect fabric with 800 Gbps per-link bandwidth and sub-microsecond latency, enabling truly distributed training across our entire compute fleet.
Exabyte-scale distributed storage system spanning three geographic regions, purpose-built for training datasets, checkpoints, and inference caches with redundant 400 Gbps uplinks.
Direct-to-chip liquid cooling and on-site 150 MW solar farm power our facility at PUE 1.08, making CNNRC one of the most energy-efficient compute centres in operation.
To advance the science of intelligence by building neural networks that learn, reason, and generalise with the efficiency and adaptability of biological systems. We believe the path to machine intelligence runs through a deeper understanding of cognition itself, and that open, collaborative research serves humanity best.
A future where intelligent systems work alongside humans as genuine collaborators, not tools. CNNRC is building the theoretical and practical foundations for machines that understand context, exhibit curiosity, and contribute to scientific discovery, creative work, and human flourishing.
Founded in Palo Alto with a seed focus on next-gen neural architectures
Broke ground on our flagship compute facility, housing 12,800 H100 nodes
Published breakthrough research on liquid neural networks in Nature AI
Expanded to three research campuses with 200+ fellows worldwide
Interested in collaborating, joining our team, or learning more about our research? We would like to hear from you.
4400 El Camino Real
Palo Alto, CA 94306
United States
research@cortex.network
+1 (650) 555-0199
Mon - Fri, 9:00 AM - 6:00 PM PST
© 2026 Cortex Network - All rights reserved