BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
Frangipani
Top 20
Mike Davies and his neuromorphic computing team at Intel Labs are not standing still either. They, too, continue to innovate and are busy developing Loihi 3 (or whatever it will eventually be called):
www.linkedin.com
sites.google.com
All the recent breakthroughs in generative AI provide great optimism that longstanding algorithmic challenges in robotics, autonomy, and embodied AI will soon be solved, but there’s one catch:… | Mike Davies
All the recent breakthroughs in generative AI provide great optimism that longstanding algorithmic challenges in robotics, autonomy, and embodied AI will soon be solved, but there’s one catch: conventional computer architectures stand no chance of delivering the TOPS needed to fulfill this...
RoboARCH @ ICRA 2025 - Schedule
Last edited:
BrainShit
Regular
Innatera just unveiled Pulsar – the world’s first mass-market neuromorphic microcontroller for real-time intelligence at the sensor edge.
And yes, it is commercially available now…
View attachment 84983
View attachment 84984
Above images are from the Press Kit: https://www.dropbox.com/scl/fo/xrpycuii9repikowerctl/AGZqj0VP0NLSuxST9CNY3Ok?
Innatera Unveils Pulsar: The World's First Mass-Market Neuromorphic Microcontroller for the Sensor Edge
/PRNewswire/ -- Innatera, a leading developer of neuromorphic processors, today announced the launch of Pulsar, its first commercially available...www.prnewswire.com
Innatera Unveils Pulsar: The World's First Mass-Market Neuromorphic Microcontroller for the Sensor Edge
NEWS PROVIDED BY
Innatera
May 21, 2025, 08:59 ET
Smaller, smarter, and radically efficient - bringing brain-inspired intelligence to battery-powered devices, unlocking a new era of real-time, ultra-low power AI at the edge.
DELFT, Netherlands, May 21, 2025 /PRNewswire/ --
Innatera, a leading developer of neuromorphic processors, today announced the launch of Pulsar, its first commercially available microcontroller to bring brain-like intelligence into edge devices. Born from more than a decade of pioneering research, Pulsar delivers up to 100X lower latency and 500X lower energy consumption than conventional AI processors. With this breakthrough, Innatera brings a new class of ultra-efficient, brain-inspired intelligence directly to the sensor edge.
With sensors embedded in everything from wearables and smart homes to cars and industrial systems, the need for real-time, secure, energy-efficient data processing at the edge has never been greater. Pulsar tackles this challenge head-on by processing data locally and intelligently, at the sensor level – eliminating the need to rely on brute-force compute in power-hungry edge processors or data centers to make sense of sensor data.
"Pulsar is not just another AI chip – it represents a fundamental shift in how we bring intelligence to the edge," says Sumeet Kumar, co-founder and CEO of Innatera. "This launch is the culmination of over a decade of deep research and engineering in neuromorphic computing, combined with a groundbreaking heterogeneous architecture. It marks the moment that our brain-inspired technology becomes ready for mass-market deployment. As demand for real-time, power-efficient intelligence in edge devices continues to grow, Pulsar delivers the capabilities that traditional AI hardware simply can't – ultra-low latency, minimal power draw, and on-device decision-making. More importantly, it lays the foundation for a new class of intelligent systems that are adaptive, autonomous, and scalable. Pulsar is our first major step toward making that future a reality."
Built for what's next: A platform for scalable, real-world edge intelligence
Pulsar introduces a compute architecture based on Spiking Neural Networks (SNNs), a generational leap in AI hardware that processes data the way the brain does, focusing only on changes in input. This event-driven model dramatically reduces energy use and latency while delivering precise, real-time decision-making. Pulsar goes even further by combining neuromorphic compute with traditional signal processing in a revolutionary architecture. Integrating a high-performance RISC-V CPU and dedicated accelerators for Convolutional Neural Networks (CNNs) and Fast Fourier Transform (FFT), this architecture provides exceptional versatility on a single chip.
"Innatera's Pulsar chip has the potential to redefine what's possible at the edge," says David Harold, senior analyst, Jon Peddie Research. "By using brain-inspired Spiking Neural Networks, it brings real-time processing to ultra-low-power devices without leaning on the cloud. That means sensors that can think for themselves – faster responses, lower energy use, and smarter performance across everything from wearables to industrial systems."
Smarter products, longer battery life
Pulsar gives product teams a shortcut to smarter features that were previously off-limits due to size, power, or complexity. Filtering and interpreting sensor data locally keeps the main application processor asleep until truly needed, in some cases, eliminating the need for a main application processor or cloud computing, extending battery life by orders of magnitude. With sub-milliwatt power consumption, Pulsar makes always-on intelligence truly viable, enabling everything from sub-millisecond gesture recognition in wearables to energy-efficient object detection in smart home systems. For example, it achieves real-time responsiveness with power budgets as low as 600 µW for radar-based presence detection and 400 µW for audio scene classification.
"The combination of Innatera's Spiking Neural Processor (SNP) and Socionext's highly integrated, sophisticated radar sensor technology introduces a powerful new approach to reducing power consumption and minimizing false detections in challenging applications, such as battery-powered devices," says Matthias Neumann, Senior Marketing Manager Smart Sensor & Smart Devices at Socionext. "We are confident that this collaboration will accelerate the adoption of radar sensing solutions in the market, bringing cutting-edge technology to a wider range of industries."
Simpler integration for sensor makers
Pulsar transforms traditional sensors into self-contained intelligent systems. With its small memory footprint and efficient neural models, it fits into tight form factors while eliminating the need for heavy external compute and reducing reliance on complex, custom DSP pipelines. Sensor manufacturers can now deliver plug-and-play smart sensor modules, accelerating development and time to market.
"Aria Sensing is committed to providing advanced, highly intuitive Ultra-Wideband system-on-chip and complete solutions. The Pulsar microcontroller by Innatera facilitates real-time sensing with exceptional energy efficiency, thereby creating opportunities for continuous operation applications," says Alessio Cacciatori, Founder and CEO of Aria Sensing. "We are particularly enthusiastic about the potential of Pulsar's neuromorphic architecture to integrate brain-inspired intelligence into our state-of-the-art 1D/2D/3D high-resolution sensing systems, leading to enhanced speed, responsiveness, and significantly improved power efficiency."
Empowering developers with familiar tools and a new community
Innatera's Talamo SDK makes neuromorphic development approachable. Developers can build spiking models from scratch, in a PyTorch-based environment, simulate, optimize, and deploy with ease.
To further support this ecosystem, Innatera is launching its developer program, now open to early adopters. More than just a portal, it's the foundation of a growing community designed to accelerate innovation, share knowledge, and empower members to build the next generation of intelligent edge applications together. An upcoming open-source PyTorch frontend and marketplace will create an even more collaborative ecosystem for neuromorphic AI.
"Innatera's Pulsar marks a leap forward in edge intelligence. As a partner, we see their neuromorphic approach redefining what's possible in ultra-low-power, always-on AI – exactly the kind of innovation the edge AI ecosystem needs to thrive," says Pete Bernard, CEO, EDGE AI FOUNDATION.
"As a trusted solution partner for innovative semiconductor technologies, SmartSoC is proud to collaborate with Innatera to bring cutting-edge AI solutions like Pulsar to market. Pulsar's unique neuromorphic architecture perfectly complements our mission to deliver smarter, more efficient products to customers across Europe and India, enabling a new generation of intelligent edge applications," says Bharath Desareddy, CEO, SmartSoC Solutions.
Designed to unlock the future
Beyond what Pulsar delivers today, it lays the groundwork for what's next: edge AI systems that are autonomous, adaptive, and capable of learning in the field. With each product generation, Innatera's roadmap leads toward self-calibrating, self-optimizing devices that reduce maintenance costs and unlock entirely new classes of edge applications.
Pulsar is available now. Whether you're building the next breakthrough in wearables, enabling smarter industrial systems, or designing cutting-edge sensors, Pulsar is your gateway to the future of edge intelligence.
Learn more and get started at www.innatera.com/pulsar.
For the press kit, click here.
About Innatera
Innatera is a pioneering semiconductor company at the forefront of neuromorphic computing, a revolutionary approach to AI that emulates the brain's neural networks. Its ultra-low-power processors are designed to deliver real-time, high-performance AI inference for energy-constrained applications. Founded in 2018 and headquartered in Delft, Netherlands, Innatera is backed by leading investors and collaborates with global industry leaders to redefine the future of edge AI. The company's mission is to bring cognitive capabilities to devices, enabling smarter, faster, and more efficient decision-making directly at the sensor.
For more information, visit www.innatera.com.
Media Contact:
Cosmin Balan
Marketing Manager
395572@email4pr.com
215-764-7336
SOURCE Innatera
Video of Innatera’s Product Launch of Pulsar earlier today at Computex 2025:
https://lnkd.in/gMgBS2AR
View attachment 84977 View attachment 84978 View attachment 84979 View attachment 84980 View attachment 84981 View attachment 84982
Innatera is indeed a very good competitor and is frequently listed alongside BrainChip, but the market is still evolving.
We'll find our place... hopefully in mass productions not only in some space/defence devices. But space/defence is the quality check ;-)
DingoBorat
Slim
I'm with you BrainShit, the competition is healthy.
Like you said, we know from our Space and Military engagements, that we have a "Premier product" in the AKIDA suite (including TENNs).
The maturing of the Edge market and of our competition, might be just what's needed, to push some deals over the line..
Sean has said, he envisioned us as ending up being one of 2 or 3, of the "Major" players in this space.
Will Innatera be one?..
Also looks like "LoiHi 3" will be commercial, but I don't think LoiHi 2 is much chop, even compared to AKIDA 1.0.. (they will have the superior marketability of their "Heritage" though, even if it "is" tainted..)
IBM, took a different direction with NorthPole..
Then there's KAIST's efforts..
Many more of course..
I wonder how CyberSwarm is doing?..
Ahh it's okay folks, Freddo will save the day
Last edited:
BrainShit
Regular
Energy efficiency analysis of Spiking Neural Networks for space applications.
The following paper work was founded by the European Space Agency (contract number: 4000135881/21/NL/GLC/my) in the framework of the Ariadna research program. An investigation of the potential benefits of Spiking Neural Networks for onboard AI applications in space was carried out in this work.
Overall, the confirmed superior energy efficiency of Spiking Neural Networks
is of extreme interest for applications limited in terms of power and energy, which is typical of the space environment, and SNN are a competitive candidate for achieving autonomy in space systems.
Source: https://arxiv.org/pdf/2505.11418
The following paper work was founded by the European Space Agency (contract number: 4000135881/21/NL/GLC/my) in the framework of the Ariadna research program. An investigation of the potential benefits of Spiking Neural Networks for onboard AI applications in space was carried out in this work.
Overall, the confirmed superior energy efficiency of Spiking Neural Networks
is of extreme interest for applications limited in terms of power and energy, which is typical of the space environment, and SNN are a competitive candidate for achieving autonomy in space systems.
Source: https://arxiv.org/pdf/2505.11418
Attachments
Bravo
If ARM was an arm, BRN would be its biceps💪!
Courtesy of ChatGPT 4
BrainChip’s Akida neuromorphic processor is designed for ultra-low power, edge AI processing, which makes it well-suited for space-based and aerospace systems, where traditional compute is often too power-hungry or slow.
Here are key areas where BrainChip could fit in — and the scale of those markets:
If BrainChip successfully embeds into:
But — big caveat — this depends on:
Estimated Annual Revenue Potential for BrainChip in Space Sector (in $M USD)
Here's a breakdown of BrainChip's estimated annual revenue potential in the space sector, assuming successful adoption:
BrainChip + Space Applications: Revenue Potential Overview
BrainChip’s Akida neuromorphic processor is designed for ultra-low power, edge AI processing, which makes it well-suited for space-based and aerospace systems, where traditional compute is often too power-hungry or slow.Here are key areas where BrainChip could fit in — and the scale of those markets:
1. Satellite Edge AI (Earth Observation, Surveillance, Comms)
- What Akida does: Onboard real-time processing of images, sensor data, or radar signals — reducing the need to transmit huge data payloads back to Earth.
- Market value:
- Global satellite data market: $12–15B annually by 2030
- Edge computing in space: Emerging but growing fast — projected $2–4B+ by 2032
2. Defense & Aerospace Systems
- Use case: Drone navigation, electronic warfare systems, secure communications, and early-warning systems — all moving toward on-board AI.
- Relevant contracts:
- US DoD spends $10B+/year on AI/ML integration
- Programs like DARPA’s “MOSAIC Warfare” and “OFFSET” involve embedded AI — including neuromorphic computing R&D.
3. Space Exploration Missions (NASA, ESA, etc.)
- Use in rovers, probes, deep space missions where latency and bandwidth make real-time Earth-based computing impossible.
- Example: Mars rovers and future lunar habitats need autonomous decision-making (e.g. image classification, navigation, anomaly detection).
- NASA and ESA have budgets over $25B/year, with increasing allocation to autonomy and onboard AI.
4. Neuromorphic AI in Space R&D (Experimental Stage)
- Agencies are experimenting with brain-inspired chips due to power efficiency.
- NASA Ames and AFRL (US Air Force Research Lab) are already trialing neuromorphic systems like IBM’s TrueNorth.
So What’s BrainChip’s Realistic Potential?
If BrainChip successfully embeds into:- Even 5–10 niche aerospace or defense systems, and/or
- One or more LEO satellite platforms, and/or
- A NASA/ESA exploratory mission,
But — big caveat — this depends on:
- Engineering validation cycles (which take years in aerospace/defense)
- Getting into partner ecosystems
- Meeting radiation-hardened chip standards (a big barrier)
Estimated Annual Revenue Potential for BrainChip in Space Sector (in $M USD)
Here's a breakdown of BrainChip's estimated annual revenue potential in the space sector, assuming successful adoption:
- LEO Satellites & Earth Observation: ~$40M
- Defense & Aerospace Systems: ~$30M
- NASA/ESA Exploration Missions: ~$20M
- Experimental R&D Contracts: ~$10M
equanimous
Norse clairvoyant shapeshifter goddess
Approval from space and defense sectors represents the pinnacle of recognition.
Here are some examples of space and military products that became mass commercial items:
- GPS (Global Positioning System)
- Origin: Developed by the U.S. military in the 1970s for navigation and positioning.
- Commercial Use: Now ubiquitous in smartphones, cars, fitness trackers, and navigation apps like Google Maps.
- Microwave Oven
- Origin: Invented accidentally during World War II when radar technology (magnetron) was found to heat food.
- Commercial Use: A common household appliance for cooking and reheating.
- Duct Tape
- Origin: Developed during World War II as a waterproof sealing tape for ammunition cases.
- Commercial Use: Widely used for repairs, construction, and DIY projects.
- Freeze-Dried Food
- Origin: Developed for space missions to provide lightweight, long-lasting food for astronauts.
- Commercial Use: Popular in camping, emergency preparedness, and instant meals.
- Memory Foam
- Origin: Created by NASA in the 1960s to improve crash protection in aircraft seats.
- Commercial Use: Used in mattresses, pillows, and furniture for comfort.
- Digital Cameras
- Origin: Early digital imaging technology was developed for spy satellites and space exploration.
- Commercial Use: Found in smartphones, cameras, and webcams.
- Cordless Power Tools
- Origin: NASA partnered with Black & Decker to develop lightweight, battery-powered tools for Apollo missions.
- Commercial Use: Common in construction, home improvement, and DIY projects.
- Smoke Detectors
- Origin: NASA developed early versions for Skylab to detect toxic gases.
- Commercial Use: Standard safety devices in homes and buildings.
- Scratch-Resistant Lenses
- Origin: NASA developed coatings for astronaut helmet visors to resist scratches and UV damage.
- Commercial Use: Used in eyeglasses, sunglasses, and camera lenses.
- Insulated Clothing
- Origin: Space suit technology inspired lightweight, heat-reflective materials for astronauts.
- Commercial Use: Found in winter jackets, sleeping bags, and outdoor gear.
These are very impressive numbers, and it would be great if one day the forecast became a flourishing reality.Courtesy of ChatGPT 4
BrainChip’s Akida neuromorphic processor is designed for ultra-low power, edge AI processing, which makes it well-suited for space-based and aerospace systems, where traditional compute is often too power-hungry or slow.
Here are key areas where BrainChip could fit in — and the scale of those markets:
Even a small sliver (e.g. AI processors inside CubeSats, LEO constellations) could yield $10M–$50M/year potential for a player like BrainChip if they secure OEM/partner deals with satellite manufacturers or defense primes.
- What Akida does: Onboard real-time processing of images, sensor data, or radar signals — reducing the need to transmit huge data payloads back to Earth.
- Market value:
- Global satellite data market: $12–15B annually by 2030
- Edge computing in space: Emerging but growing fast — projected $2–4B+ by 2032
If BrainChip becomes part of a standard subsystem or receives a subcontract through a partner (like Lockheed or Northrop), even single programs could generate $5M–$20M+ per project.
- Use case: Drone navigation, electronic warfare systems, secure communications, and early-warning systems — all moving toward on-board AI.
- Relevant contracts:
- US DoD spends $10B+/year on AI/ML integration
- Programs like DARPA’s “MOSAIC Warfare” and “OFFSET” involve embedded AI — including neuromorphic computing R&D.
If Akida lands in even one long-term exploratory platform, it could bring low tens of millions in royalties or contract revenue over the mission lifecycle.
- Use in rovers, probes, deep space missions where latency and bandwidth make real-time Earth-based computing impossible.
- Example: Mars rovers and future lunar habitats need autonomous decision-making (e.g. image classification, navigation, anomaly detection).
- NASA and ESA have budgets over $25B/year, with increasing allocation to autonomy and onboard AI.
Though early-stage, R&D contracts and prototype testbeds can range $500K to $5M per deal — and are a key stepping stone to deployment.
- Agencies are experimenting with brain-inspired chips due to power efficiency.
- NASA Ames and AFRL (US Air Force Research Lab) are already trialing neuromorphic systems like IBM’s TrueNorth.
If BrainChip successfully embeds into:
then it could realistically unlock $20M–$100M+ annual revenue within 3–5 years, with longer-term licensing potential into the hundreds of millions if neuromorphic tech becomes standard in edge-AI space processing.
- Even 5–10 niche aerospace or defense systems, and/or
- One or more LEO satellite platforms, and/or
- A NASA/ESA exploratory mission,
But — big caveat — this depends on:
- Engineering validation cycles (which take years in aerospace/defense)
- Getting into partner ecosystems
- Meeting radiation-hardened chip standards (a big barrier)
Estimated Annual Revenue Potential for BrainChip in Space Sector (in $M USD)
Here's a breakdown of BrainChip's estimated annual revenue potential in the space sector, assuming successful adoption:
These figures represent a realistic upside scenario within 3–5 years if neuromorphic computing gains traction in key programs. Let me know if you’d like a 10-year projection or a version based on conservative, moderate, and aggressive adoption cases.
- LEO Satellites & Earth Observation: ~$40M
- Defense & Aerospace Systems: ~$30M
- NASA/ESA Exploration Missions: ~$20M
- Experimental R&D Contracts: ~$10M
However, I now doubt whether Brainchip management is smart enough to make the projected sales a reality.
I could rather imagine someone making a suggestion to Brainchip management:
"If you provide us with Akida free of charge, then Brainchip won't have to pay for the great advertising that Akida is in Mars Rovers and satellites."
equanimous
Norse clairvoyant shapeshifter goddess
This is a very impressive post.
If I was in your shoes and had this much doubt I would sell and move on.
Correct me if I am wrong but I think Innatera T1 doe not have on chip learning as AKIDA does.
Impressive numbers. Cannot comment on achievabiliy.
However AKIDA onchip learning is the key enabler of cybersecurity which will be important for Health,space, defence, Auto ect
At this stage not that many chips have on chip learning and also offer the broad range of uses that AKIDA does.
DingoBorat
Slim
Good presentation.
Sean says from the 3.48min mark, that we will be hitting our stride, within the next 10 to 20 years.
Thanks for posting. Great presentation.Good presentation.
Sean says from the 3.48min mark, that we will be hitting our stride, within the next 10 to 20 years.
Sean preparing to retain Edge industry leadership as the Edge grows.
Sean confirmed that there are auto engagements during question time.
Labsy
Regular
I'd be happy with those figures....Courtesy of ChatGPT 4
BrainChip’s Akida neuromorphic processor is designed for ultra-low power, edge AI processing, which makes it well-suited for space-based and aerospace systems, where traditional compute is often too power-hungry or slow.
Here are key areas where BrainChip could fit in — and the scale of those markets:
Even a small sliver (e.g. AI processors inside CubeSats, LEO constellations) could yield $10M–$50M/year potential for a player like BrainChip if they secure OEM/partner deals with satellite manufacturers or defense primes.
- What Akida does: Onboard real-time processing of images, sensor data, or radar signals — reducing the need to transmit huge data payloads back to Earth.
- Market value:
- Global satellite data market: $12–15B annually by 2030
- Edge computing in space: Emerging but growing fast — projected $2–4B+ by 2032
If BrainChip becomes part of a standard subsystem or receives a subcontract through a partner (like Lockheed or Northrop), even single programs could generate $5M–$20M+ per project.
- Use case: Drone navigation, electronic warfare systems, secure communications, and early-warning systems — all moving toward on-board AI.
- Relevant contracts:
- US DoD spends $10B+/year on AI/ML integration
- Programs like DARPA’s “MOSAIC Warfare” and “OFFSET” involve embedded AI — including neuromorphic computing R&D.
If Akida lands in even one long-term exploratory platform, it could bring low tens of millions in royalties or contract revenue over the mission lifecycle.
- Use in rovers, probes, deep space missions where latency and bandwidth make real-time Earth-based computing impossible.
- Example: Mars rovers and future lunar habitats need autonomous decision-making (e.g. image classification, navigation, anomaly detection).
- NASA and ESA have budgets over $25B/year, with increasing allocation to autonomy and onboard AI.
Though early-stage, R&D contracts and prototype testbeds can range $500K to $5M per deal — and are a key stepping stone to deployment.
- Agencies are experimenting with brain-inspired chips due to power efficiency.
- NASA Ames and AFRL (US Air Force Research Lab) are already trialing neuromorphic systems like IBM’s TrueNorth.
If BrainChip successfully embeds into:
then it could realistically unlock $20M–$100M+ annual revenue within 3–5 years, with longer-term licensing potential into the hundreds of millions if neuromorphic tech becomes standard in edge-AI space processing.
- Even 5–10 niche aerospace or defense systems, and/or
- One or more LEO satellite platforms, and/or
- A NASA/ESA exploratory mission,
But — big caveat — this depends on:
- Engineering validation cycles (which take years in aerospace/defense)
- Getting into partner ecosystems
- Meeting radiation-hardened chip standards (a big barrier)
Estimated Annual Revenue Potential for BrainChip in Space Sector (in $M USD)
Here's a breakdown of BrainChip's estimated annual revenue potential in the space sector, assuming successful adoption:
These figures represent a realistic upside scenario within 3–5 years if neuromorphic computing gains traction in key programs. Let me know if you’d like a 10-year projection or a version based on conservative, moderate, and aggressive adoption cases.
- LEO Satellites & Earth Observation: ~$40M
- Defense & Aerospace Systems: ~$30M
- NASA/ESA Exploration Missions: ~$20M
- Experimental R&D Contracts: ~$10M
Also…Sean confirmed the company that is manufacturing the glasses for Onsor is the same one doing them for Meta. I thought that was interesting synergy there.
Labsy
Regular
Maybe you should stop laughing and hear what he says at 14.29Good presentation.
Sean says from the 3.48min mark, that we will be hitting our stride, within the next 10 to 20 years.
Onsor" moving aggressively and frames are being designed by none other that LUXOTTICA! The same company (Ray bans) who are making the meta glasses. Based on this little nugget I'm gonna buy a shit ton more shares in the next few weeks...
Now back to your whining, crying and laughing.
PS I think Sean presented amazing well here. Im a fan....
Similar threads
