Birmingham, UK – November 2025
Qognetix has been selected to present at the Spiking Neural Universal Function Approximation (SNUFA) 2025 meeting — a leading international forum on spiking neural computation and biologically grounded learning systems.
The presentation, titled “Demonstrating Universal Approximation with a Biophysically Faithful Hodgkin–Huxley Engine,” demonstrates how Qognetix’s biophysically faithful Engine achieves real-time Hodgkin–Huxley computation on standard hardware — bringing the power of detailed neural models out of supercomputing labs and into accessible environments.
“Most neuromorphic and AI systems simplify neurons to make them scale,” said Nic Windley, co-founder of Qognetix. “Our work shows you can have both — full biological fidelity and computational efficiency — proving that physics itself can compute.”
The session highlights include:
- Benchmarks against major simulation platforms such as NEURON, NEST, and Brian2.
- Demonstrations of universal function approximation and temporal learning using full Hodgkin–Huxley neurons.
- A preview of BioSynapStudio, Qognetix’s integrated environment for designing and training HH-class experiments without low-level coding.
- An introduction to the Synthetic Endocrine Controller (SEC) Layer, a hormone-inspired feedback system for adaptive stability and emotion-modulated learning.
SNUFA 2025 is hosted virtually by an international programme committee including Dr Julijana Gjorgjieva, Dr David Kappel, Dr Dan Goodman, and Dr Friedemann Zenke. Qognetix was selected for a poster presentation among a highly competitive group of computational neuroscience researchers.
“We see this as an important milestone for Qognetix — recognition that our approach bridges neuroscience and computation,” added Windley. “It validates our core thesis: real intelligence starts with real neurons.”
For more information on the research or to view the presentation abstract, visit: Demonstrating Universal Approximation with a Biophysically Faithful Hodgkin–Huxley Engine | SNUFA.
Missed our presentation ? Catch up with it here > SNUFA 2025 – Qognetix
About Qognetix
Qognetix is a UK-based deep-tech venture developing biologically grounded Synthetic Intelligence. Its flagship platform, BioSynapStudio, combines biophysical accuracy with computational efficiency, enabling realistic spiking neural simulation and emotion-modulated learning on commodity hardware.
