Long post ahead - if you’re not interested in what I found out about Steve Harbour’s invention The Living Processor™, whose concept fundamentally differs from fully digital neuromorphic chips such as Loihi and Akida, simply scroll on to the Steve Harbour quote marked in
green - I am pretty sure you will enjoy reading that part…
Steve Harbour - formerly with SwRI (Southwest Research Institute) in Beavercreek, OH (a suburb of Dayton) and since June 2024 Director of AI Hardware Research at Parallax Advanced Research, also in Beavercreek, OH - has been working with Loihi for years, but has also been utilising Akida in his neuromorphic research, at least since joining Parallax. He has repeatedly spoken favourably about both Loihi and Akida and revealed in an article published on 29 April 2025 ("Parallax Advanced Research and the Future of Neuromorphic Artificial Intelligence in Electronic Warfare")
that Parallax were partnered with both Intel and BrainChip, something that to date is still not reflected on the BrainChip "Partners" webpage.
A paper on Martian flight [= flight of a mini helicopter/drone in the extremely thin atmosphere of Mars] published in July 2025* that he co-authored with his son David Harbour (first author) and other researchers from Parallax Advanced Research, University of Dayton, University of Cincinnati and Sinclair College mentions that the neuromorphic system they developed - dubbed VelocitySNN-Fuzzy AI - “supports autonomous flight with interpretable output and deployment on neuromorphic hardware such as Loihi and Akida”. Funding for this research came from NASA.
*“Event-Driven Spiking Neural Network and Fuzzy Logic AI System for Velocity Determination in Martian Flight”
View attachment 94127
View attachment 94128
However, over the past few months it has become increasingly clear that fully digital neuromorphic processors like Loihi and Akida are not the holy grail in Steve Harbour’s eyes.
In October, I shared a podcast titled "Thinking Like the Brain: Neuromorphic AI and the Future of Defense Computing”, in which he was interviewed by Arun Seraphin, the Executive Director of the NDIA Emerging Technologies Institute (ETI) and a former Professional Staff Member on the staff of the United States Senate Committee on Armed Services:
At around the 23:30 min mark, Arun Seraphin asks "How mature is the technology, you know, on a TRL scale? Can I go to Costco and buy a laptop with a neuromorphic processor on it, yet, or Best Buy? Or is it something that I see at someone's lab bench, and it's all sort of all wired up right now, and they're promising me in five years?"
TRL= Technology Readiness Level
Steve Harbour replies that “it’s in a stage that I think we are beyond just a lab curiosity” and then briefly talks about both Intel and BrainChip, but also shares his belief that the future of neuromorphic chips will ultimately belong to analog neuromorphic processors, adding that he is currently working on a novel processor that will be largely analog:
From 24:56 min
“... BrainChip Akida - probably has, if not the first commercialised neuromorphic processor, it's among the first commercialised. And you can buy that today. And then you get support with it.
Again, it’s a digital-based neuromorphic processor with a neuromorphic mesh, if you will. Full-blown neuromorphics will involve something that's very analog-like, okay, like the mind. And we’re dev, we're working on that. In fact I’m heavily involved in a specific program right now doing that.”
At 31:33 min, he reiterates this assessment:
"And at the end of the day we're gonna see, after the long haul, that the analog neuromorphic-like processor is going to win out."
From 42:33 min, Steve Harbour reveals a little bit more about the Living Processor project he alluded to minutes earlier:
"I think besides the military on the edge, in space on the edge, you know, I think we're gonna see it with wearable devices in the medical community, whether they are on or inside the human. I think also, I think the data centers are going to be a huge driver. One of the, so, erm,
I've been an innovator for something that Parallax called the Living Processor, and I'm working on that with some universities on doing research. And it's a processor that could be used for the edge or for the data center.
And so, I think that that's gonna be - I don't know which is gonna, you know, we know that in the military, it's gonna be the edge device, it's just gonna happen. In space, it's gonna be the device of choice. I believe that it's gonna be, I don't know which is gonna happen first, either the, you know, the smartphone with a neuromorphic processor or if it's gonna be the data center. You know, there's gonna be, again, some challenges with the data center, you know, it's a little different way of computing, with LLMs, but, you know, I know we can breakthrough that one, too. So I think it's gonna be, who's gonna be first in the civilian market? Is it gonna be cell phone or data center? But all that's gonna find its way in the personal computer, and you know, and become more the norm."
At 48:49 min, the podcast host asks
"Is there work going on between this Neuromorphic Computing AI community and those people who really do study legitimately the brain - to develop more optimal systems?"
SH: "...
NIH is a pivotal piece of this process. You know, for example, I work with Dr. Rishma [he means Rashmi] Jha at UC [= University of Cincinnati] and folks at Penn State.
We currently have an NSF Grant for a DNA compute layer - I said DNA compute layer - to add to work that we're currently doing with this Living Processor."
Steve Harbour also mentions being the “Inventor of the Living Processor™” in his LinkedIn profile:
National security AI and semiconductor R&D leader with 26+ years advancing neuromorphic… · Experience: Parallax Advanced Research · Education: Massachusetts Institute of Technology · Location: Dayton · 500+ connections on LinkedIn. View Dr. Steve Harbour, PhD’s profile on LinkedIn, a...
www.linkedin.com
“
Innovation in National Security AI: Invented the Living Microprocessor™ — a hybrid neuromorphic architecture enabling self-healing, ultra-low power, and event-driven computing directly relevant to DOE/LANL’s scalable secure AI mission.”
LANL = Los Alamos National Laboratory (https://www.lanl.gov)
On 26 September 2025, Parallax Advanced Research published a blog post called "Toward a Living Microprocessor™: Dr. Steve Harbour’s Vision for Bio-Integrated Intelligence"
https://parallaxresearch.org/news/b...e-harbours-vision-bio-integrated-intelligence
“The Living Microprocessor™ integrates material innovations, both organic and inorganic, into a single architecture. Together, these materials create a processor that functions like a cognitive organism with near-zero latency and ultra-low power consumption. Unlike digital neuromorphic platforms, Harbour’s design introduces organic intelligence into the hardware itself. In practical terms, this means a processor that not only maintains function under noise, radiation, or physical damage but actually improves in adverse conditions.
The implications extend far beyond laboratory performance. Harbour envisions these processors as the backbone of next-generation human-machine teaming.
This innovation could also apply to RF classification and improve data center scalability. Because the Living Processor is robust to radiation and electromagnetic interference, it is also ideally suited for space and electronic warfare environments. The Living Processor is a collaborative effort among Parallax, the University of Dayton, the University of Cincinnati, Rochester Institute of Technology, Pennsylvania State University, and Brisk Computing.”
Brisk Computing LLC (
https://briskcomputing.com), which has won four Phase I SBIR and two Phase II SBIR awards (5x NASA, 1x DoD) since 2020 that were all related to neuromorphic computing, is registered at the residential address of Tarek M. Taha, one of the University of Dayton neuromorphic researchers who has been collaborating with Steve Harbour for years:
https://www.sbir.gov/portfolio/1676181
https://www.ohioresidentdatabase.com/person/OH0020935927/taha-tarek
The same day the Parallax blog post came out, Steve Harbour gave a presentation at an event called “Cybersecurity in the era of AI”, organised by Ohio State University’s Institute for Cybersecurity & Digital Trust. A video of this presentation titled “Brains over Bots: How Bio-Inspired Computing Transforms Cyber Defense” was uploaded to YouTube on 2 October 2025.
View attachment 94129
Unfortunately, the audio quality is not very good, and there were quite a few parts that were unintelligible to me. I marked them as [x] in my quotes below. Not 100% sure I got the rest right, but as always I’d advise anyone to listen to interviews/podcasts themselves anyway rather than simply rely on what other forum members post here and elsewhere. Also, it’s about noticing intonation, emphasis, sentences being rephrased - subtle hints that may not be picked up when reading a transcript, even a verbatim one.
Steve Harbour refers to his Living Processor-work in progress in one of the presentation slides (from 10:48 min) as well as in the Q&A section. He clearly differentiates it from Loihi and Akida. Not only will the Living Processor have a hybrid architecture - partly digital, but mostly analog. Unlike Loihi or Akida, it is not a CMOS chip, but instead a mix of organic and inorganic materials, although Steve Harbour clarifies they are not using "fleshy" brain tissue as organic material.
While he refers to both Loihi and Akida as really good neuromorphic chips for edge applications, he evidently doesn’t consider them suitable for use in big data centers - something that he claims differentiates them from his own invention which he predicts will ultimately play a dual role - both at the edge and in the data centre. [I suppose Intel Labs and BrainChip would disagree with him here, as they see their own offerings as eventually scalable.]
From 19:27 min:
Question from the audience: “Are you guys manufacturing this at scale, yet, or is it still mostly [doesn’t finish the sentence]?”
SH: “Yeah, not at scale, yet.
So, erm, I've done a lot of work with Mike Davies and Intel, San Jose in [and?] neuromorphic processors, Loihi 1, Loihi 2, and they're real good at edge - they’re research chips, you can't go buy these - they're real good at edge computing. They're not [x, possibly scalable?]. That's a problem.
Erm, BrainChip. Ever heard of BrainChip? Check them out. BrainChip, they have an AKD1000 that you can buy. They just put it in commercial availability to buy it. And they are around 500 [US]$ and you can start working with it, playing with it, okay. It's not a full neuromorphic chip, it's still digital. It has a neuromorphic mesh, so it behaves kinda [x], but it uses a lot less energy and it has very good accuracy when trained, you know when [x] final result [x] deploy the algorithm, but again, it's edge, [x, possibly fit or for?] the edge, which is really good.
So, the Living Processor that I put up there, that one will be, can be dual role - at the edge, but then also in the data center. The memory capability - can't say what it is, but it is, it will blow your mind, it wholly will blow your mind. And it’s, it’s with the memory, computing memory and also storage within the layer, yeah, you can replace [x] racks of GPUs in the data center.
So, how fast is that gonna happen? Is it gonna happen in a year? You know, it's gonna be a little bit of time, but we have to get there as fast as we can, because we're getting more and more data centers. Everyone’s now doing the shopping list with ChatGPT apparently. [x] “It’s a great tool. I think it’s [x, sounds like ‘dual on steroids’?]”.
Next up is the question “Are the big GPU providers that are out there already leaning in this direction? Have you seen evidence of that?”, to which Steve Harbour replies: “So, I have not seen evidence of that [goes on to briefly talk about NVIDIA]… I have no indication they are, but, you know, my sixth sense is they’re probably going ‘You know what? We don’t wanna be the dodo or the dinosaur.’ Right? Some companies have done that and then they get bad, right? So, but I’m not seeing if they are [x].
IBM has. IBM has a couple of, has NorthPole, which is, again, it’s a digitally-based neuromorphic mesh chip, in research, so, it’s, it’s, you know, starting to see that.
Raytheon has - I'm not a business guy, but I've been told that Raytheon’s bought some stock, [x] stock in BrainChip.
So, you know, it's starting to happen, momentum [x], everyone is seeing the writing on the wall, [x]."
Q: “So are these Loihi chips from Intel, are they, like, truly analog, and also if they aren't, erm...”
SH: “Yeah, the are analog, unfortunately. So, [corrects himself] they're digital.
The Living Processor is mainly analog, some digital. So for an absolute pure neuromorphic, it's analog, right? And - I know, I know: [copies other people who will say in disbelief] 'What? Is this not like going backwards? ' [...]”
At the end of the Q&A, somebody in the audience asks about the computational medium they are using and specifically whether they have considered using mushrooms (no kidding, research into fungal computing has been around for a while and mycelium as a computing substrate is said to be promising; cf this recent research into organic memristors made of shiitake and button mushrooms conducted at Ohio State, which happened to host the event:
https://news.osu.edu/powered-by-mushrooms-living-computers-are-on-the-rise/):
SH replies: "Mycelium?
So, erm, I can't tell you about the composition, I'm sorry [...]
and DARPA is involved. So. But it’s a great question, and, it’s a very, it’s a good question, by the way. One I’d ask.
Mycelium? So, the answer is, I’m doing research in mycelium […] Under our feet, the trees are talking, okay? So, everybody’s got trees, we have stuff underneath our feet, right? There’s communication going on, guys. It’s going on, okay? And it’s going on in a spiking fashion, like our brains. And there may be some language going on, too. So yes, the answer is ‘yes’ to your question. I was just [?] with DARPA very much on this specific thing. So,
that I can share.”
So at 22:50 min, Steve Harbour reveals a piece of information that is new to us BRN shareholders, when he claims: “I've been told that Raytheon bought some stock, [x] stock in BrainChip”, namely that Raytheon/RTX have apparently put their money where their mouth is.
Saying that, they are yet to sign an actual IP license.
While this stock purchase info could be just a rumour, as long as we do not have official confirmation from BrainChip, Raytheon/RTX or via publicised stock market documents, I personally believe Steve Harbour would likely have heard about it from somebody in the know, not merely on the grapevine.
So provided this investment info is accurate, it appears that Raytheon not only believes in BrainChip's technology, but also in our present or future management's ability to translate this technology into meaningful revenue at some point.
should read this paper and look at the pictures 
