Project Phasor is a collaborative open-source initiative building the shared infrastructure — compilers, virtual machines, and governance standards — that the neuromorphic ecosystem has been missing.
Neuromorphic computing holds transformative promise: brain-inspired processors capable of running intelligence with orders of magnitude less power than today's GPUs. Yet despite decades of research, real-world adoption remains frustratingly slow.
The reason is fragmentation. Academic labs build novel hardware with no path to production. Commercial vendors develop proprietary stacks that don't interoperate. Every team reinvents the same low-level infrastructure, over and over.
"Bridges are sorely needed between commercial and academic neuromorphic efforts." — Dylan Muir, SynSense
Project Phasor is building those bridges. We develop shared, open-source infrastructure — virtual machines for hardware exploration, a production compiler framework, and common governance standards — that the entire neuromorphic ecosystem can build on together.
Address the disconnect where research prioritizes novelty and commercial entities can't invest in non-proprietary standards.
Develop compilers, virtual machines, and tooling that no single organization would build alone — freely available to all.
Work with ML Commons, Edge AI Foundation, OCP, and Khronos to establish the interoperability standards the field needs.
Lower barriers so neuromorphic hardware becomes a genuine part of the global compute landscape.
Phasor's roadmap is built around three complementary efforts forming a complete neuromorphic software stack.
A curated library of behavioral and cycle-accurate reference models for neuromorphic cores and systems. Enables hardware exploration, architecture design, and debugging without physical silicon — leveraging established SoC design tools like SystemC.
SimulatorA production-quality lower-level compiler built with a "compiler-first mentality." A foundational layer optimized for heterogeneous integration — providing the flexibility for higher-level frameworks to build upon rather than replacing them.
CompilerEngaging ML Commons, Edge AI Foundation, OCP, and Khronos to establish common governance. Working groups cover funding models, benchmarking suites tailored to neuromorphic systems, and cross-platform compatibility frameworks.
StandardsStructured around three working groups — Community, Research, and Applications — led by a founding team with deep expertise across neuromorphic hardware, software, and ecosystem building.
Project Phasor has attracted endorsers and contributors from universities, national labs, startups, and industry across the globe.
Project Phasor is an open community effort. Whether you're a researcher, hardware vendor, software engineer, or institution — there's a place for you. Attend working group meetings, contribute to the codebase, add your endorsement, or partner with us.