Neuromorphic computing
Neuromorphic computing is starting to appear in sensor products. Innatera and Joya reported a Pulsar-based audio module in March, first shown classifying baby cries. A named module is better evidence than general efficiency claims. High-volume consumer shipments are still unproven, and so is any advantage in unrelated AI work.
Research snapshot · updated Nov 12, 2026
State of the art (2026)
Neuromorphic remains a research field with a thin commercial edge. The flagship system, Intel's Hala Point (1.15 billion neurons across 1,152 Loihi 2 chips), runs at Sandia as a research platform, not a product, and Intel has shipped no official Loihi 3 silicon. IBM's NorthPole is likewise a research inference architecture. The real traction is at the edge: BrainChip, the only listed pure-play, raised $25m in December 2025 to scale Akida 2; Innatera began shipping its Pulsar SNN microcontroller into consumer wearables and smart-home devices in early 2026; SpiNNcloud is commercialising SpiNNaker 2. Patent filings surged in 2025, but spiking-network software tooling and a killer high-volume application are still missing.
Core thesis
Neuromorphic chips take ideas from the brain to cut the power needed for AI tasks. Many use spiking neural networks, which compute only when an event arrives. Some do not: IBM's NorthPole is brain-inspired but moved away from spiking, and on the ResNet-50 image test it processed 25 times more frames per watt than a comparable 12nm GPU. That result is for one benchmark. Software tools for spiking chips are far behind mainstream AI tools, and most current AI models do not map well onto spikes. The nearest commercial use is in always-on sensors and event-based cameras.
Explore the deeper analysis in CanaryIQ
Explore what is covered below. These previews show the structure; log in to read the full analysis.
Signal stack
Evidence stacked leading → lagging
6 signals
Signal stack
Evidence stacked leading → lagging
Technology-native KPIs
Metrics that predict trajectory, tracked over time
5 tracked
Technology-native KPIs
Metrics that predict trajectory, tracked over time
Landscape map
Who builds what, and who depends on whom
17 players · 4 layers
Landscape map
Who builds what, and who depends on whom
Catalyst calendar
Dated events that will move the position
6 ahead
Catalyst calendar
Dated events that will move the position
Technology roadmap
Milestones on the path to maturity
9 milestones
Technology roadmap
Milestones on the path to maturity
Watchlists
Companies, people and papers, each with a condition for removing it
19 · 20
Watchlists
Companies, people and papers, each with a condition for removing it
Decision frameworks
What our view means for your job
Locked
Decision frameworks
What our view means for your job
Thesis changelog
When our view changed, and why
9 updates
Thesis changelog
When our view changed, and why
Change our mind
2 disconfirming conditions
You've read the verdict. The file is much deeper.
The full signal stack, technology-native KPIs tracked over time, the landscape of who depends on whom, the dated catalyst calendar, decision frameworks for every desk, live watchlists and a record of every time our view on Neuromorphic computing has changed. All of it is in CanaryIQ.