We spent the last 11 years re-thinking AI hardware from first principles
Now, in 2025, we're ready to make it a reality. Here are the key milestones from our journey
2014 - Project begins
Project Begins
Our project began in a research lab in 2014, when we began to explore the applications of photonics and superconductors to neuromorphic computing.
2017 - Silicon LEDs
All-silicon optical links at the single-photon level
We demonstrated emission of light at 1.2um in silicon integrated waveguides coupled to superconducting single-photon detectors, a key requirement for efficient communication in our platform.
2018 - Multiplanar Waveguides
Optical circuits that route light with multiple planes of waveguides
To take full advantage of optical communication, we need to maximize the number of independent channels that can be used to link neurons to other synapses. Here, we demonstrated up to 3 planes of waveguides on one substrate.
2018 - Superconducting Loop Neuron Concept
Superconducting Loop Neuron Concept
We published the original concept of a superconducting loop neuron, forming the basis of our technology going forward.
2019 - hTrons
A superconducting amplifier to interface superconductors to semiconductors
Superconducting impedance converters known as hTrons were demonstrated in our lab, enabling a link between low-voltage superconductors and semiconductors.
2022 - Synapses
Superconducting optoelectronic single-photon synapses
Combining Josephson junctions and superconducting detectors, we realized the first superconducting optoelectronic single-photon synapse. It responded to presynaptic spike rates of >10 MHz and consumed only 33 aJ of dynamic power per synapse event.
2023 - Phenomenological Model
Mathematical framework relating our circuits to all of AI
We developed a phenomenological model that increases simulation speeds by a factor of ten thousand, while maintaining close agreement with the brute force approach. This has since evolved into a high-level simulation codebase for modelling full networks and experimenting with neuromorphic algorithms.
2024 - Memory Cells
Co-location of memory with processing
Superconducting nanowire single-photon detectors and Josephson junctions are combined into programmable synaptic circuits that exhibit single-photon sensitivity, memory cells with more than 400 internal states, leaky integration of inputs, and 0.4 fJ programming energies (including cooling power).
2024 - GreatSky
Founding of the company
We made the leap from lab to startup, and GreatSky was formed.
2014 - Project begins
Project Begins
Our project began in a research lab in 2014, when we began to explore the applications of photonics and superconductors to neuromorphic computing.
Read more
2017 - Silicon LEDs
All-silicon light-emitting diodes waveguide-integrated with superconducting single-photon detectors
We demonstrated emission of light at 1.2um in silicon integrated waveguides at cryogenic temperatures, a key requirement for efficient communication in our platform.
Read more
2017 - Multiplanar Waveguides
Multi-planar amorphous silicon photonics with compact interplanar couplers, cross talk mitigation, and low crossing loss
To take full advantage of optical communication, we need to maximize the number of independent channels that can be used to link neurons to other synapses. Here, we demonstrated up to 3 planes of waveguides on one substrate.
Read more
2018 - Superconducting Loop Neuron Concept
Superconducting Loop Neuron Concept
We published the original concept of a superconducting loop neuron, forming the basis of our technology going forward.
Read more
2019 - hTrons
A superconducting thermal switch with ultrahigh impedance for interfacing superconductors to semiconductors
Superconducting impedance converters known as hTrons were demonstrated in our lab, enabling a link between low-voltage superconductors and semiconductors.
Read more
2022 - Synapses
Superconducting optoelectronic single-photon synapses
Combining Josephson junctions and superconducting detectors, we realized the first superconducting optoelectronic single-photon synapse. It responded to presynaptic spike rates of >10 MHz and consumed only 33 aJ of dynamic power per synapse event.
Read more
2023 - Phenomenological Model
Phenomenological model of superconducting optoelectronic loop neurons
We developed a phenomenological model that increases simulation speeds by a factor of ten thousand, while maintaining close agreement with the brute force approach. This has since evolved into a high-level simulation codebase for modelling full networks and experimenting with neuromorphic algorithms.
Read more
2024 - Memory Cells
Programmable superconducting optoelectronic single-photon synapses with integrated multi-state memory
Superconducting nanowire single-photon detectors and Josephson junctions are combined into programmable synaptic circuits that exhibit single-photon sensitivity, memory cells with more than 400 internal states, leaky integration of input spike events, and 0.4 fJ programming energies (including cooling power).
Read more
2024 - Great Sky
Phenomenological model of superconducting optoelectronic loop neurons
We made the leap from lab to startup, and Great Sky was founded.