Neuromorphic Simulation Intern (Undergraduate)
Description:
Cambrya is developing Synchrony Loop Networks (SLNs), a new evolution of Spiking Neural Networks (SNNs) capable of real-time, lifelong, and unsupervised learning for complex sensory tasks previously beyond neuromorphic capabilities. SLNs are powered by a new class of learning rule called Synchrony Loop Propagation (SLP). SLP is implemented within a heterogeneous architecture encompassing multiple subnetworks with different neuron types. A proof-of-concept SLN designed for auditory processing achieved best-in-class zero-shot sound source separation for complex auditory scenes containing 2 to 20 intermixed sound signals on a single channel, without training or supervision (listen here).
SLNs are fundamentally compatible with neuromorphic constraints, but incorporate synaptic mechanisms not yet supported by current standard neuromorphic architectures. To bridge this gap, Cambrya is partnering with Dr. James Plank at TENNLab to sponsor a summer intern to extend TENNLab’s framework-open codebase for the RISP neuroprocessor to support simulating an exemplar SLN. This project will help lay the groundwork for implementing SLNs in FPGA, while also enhancing RISP with a new and advanced suite of synaptic capabilities.
An undergraduate intern at Cambrya is responsible for working with our algorithms lead to learn core SLN architectural and computational components, extending TENNLab’s framework-open codebase to support these components, and implementing a small example SLN neuroprocessor simulation through framework-open.
Cambrya is composed of a small team, and is formed around shared values of curiosity, discovery, and stewardship. We have a culture of open communication and collaboration, with all team members encouraged to share their own expertise and seek out the expertise of others.
Reports to: Patrick Abbs, Chief of Algorithms & Strategy at Cambrya
Advisor: Dr. James Plank, Professor of EECS at the University of Tennessee
Duration: Summer 2025 Internship – 12 weeks
Working hours: 9:00 AM -5:00 PM ET
Location: Remote (Cambrya is based in Austin, TX)
Compensation: Hourly, $30/hour
Core Work:
Collaborate with Cambrya’s algorithm lead to learn SLN concepts and mechanisms
Identify SLN computational elements that lack RISP neuroprocessor support
Design and develop support for these components in the framework-open RISP simulator C++ codebase
Implement a small test SLN using the newly supported components in framework-open
Responsibilities:
Collaborate with Cambrya algorithm developers and business leads to scope, build, test, and deliver core work
Technical ownership of framework-open SLN component code design and development of exemplar simulation prototype in consultation with business and product leads
Develop a working understanding of SLN algorithm architecture and function
Key Skill Sets:
Demonstrated ability to develop in C++
Familiarity with RISP framework-open codebase
Working knowledge of standard integrate-and-fire neurons
Ability to work independently and as a part of a team
Ability to distill complex information into concise and interpretable form
Experience with managing code releases, branching, and merging a plus
Some familiarity with C# (current SLN codebase language) a plus
To apply, send a current CV and cover letter to careers@cambrya.co