Pom down under
Top 20
Well we know you don’t, isn’t that all that counts
BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
My dream to this day has never changed, that is to see both Peter and Anil receive the true acknowledgement that they both deserve.
I agree that key staff should always be given incentives, but the timing of such is also critical, we have not only burnt funds over the years, but since the days when Peter took charge, the outgoings have stablized to around 4 million US per quarter and seems pretty steady and controlled in my opinion.
Sean has done positive improvements in my opinion, the second half of 2026 through to mid 2027 is the period I am expecting to see rewards for all the hard work behind the scenes.
Cheers Tech.
Jacob Kiraz
Regular
Yes reflecting over the weekend Its about Time,
In saying that the next 12 months is key to see the company with more engagements and possibly some IP deals, it would be good if we had some solid information with the future of megachips and Renesa whether there in or out,
When will the AI world really kick into gear,
The AGM I feel some hard hitting questions will be presented and hopefully answered,
Especially to Sean to explain the comments over the last few months
Last edited:
Hi Tech,
Sean's stated aim is that longer term Brainchip will be in the top 2 or 3 leaders of a much larger and growing Neuromorphic Edge AI Industry.
This means that Brainchip is well aware that competition will emerge.
It follows that in order to stay on top of 'our own game' we need to keep developing and improving our tech and ecosystem non stop even before the frustrating wait until deals happen - and IMO they will.
Staff developing tech and the ecosystem need to be rewarded.
Sean reiterated in the update that there were engagements underway. I imagine staff initiating these would at some stage see rewards.
The employee share scheme holds a lot of shares some of which end up with staff.
Hi Dio, i remember Tony Lewis said with Gen 2 a bright student could invent something at home.
Kevin D. Johnson of IBM bought an AKIDA 1000 M.2.
It's interesting that he can apparently just plug into Symphony on a Monday and off it goes.
I am aware he is practically an AI genius but i wonder whether he would have 'quietly' needed some prior advice from BRN?
The way its come out on his linked in post makes it sound so easy and is great marketing for both IBM and Brainchip.
Especially great marketing for Brainchip as it cannot be seen as company ramping.
I think that IBM will be very good for Brainchip deal wise.
Hi dodgy. You are next level. I do hope you are earning some seriously big $$ as your intellect and technical understanding is next level and way above my pay grade and probably many others here. Really appreciate your contribution and technical breakdown of the tech. Much appreciated and respected here.
Hi manny,
To get it running, he would have needed an Akida compatible model and then he would have needed to configure the layers/NPUs per layer/inter-NPU connexions/weights per connexion.
The model would have been adapted from Symphony's model to Akida 4-bit format using MetaTF.
I think there is also guidance for the configuration.
No doubt his AI/NN expertise would have made it a piece of cake.
Akida and MetaTF were intended to be implementable by a competent AI/NN programmer. Familiarity with PyTorch/Keras would be an advantage.
The big hurdle that any newcomer faces is obtaining an appropriate model. BRN provides basic models which can be expanded with on-chip learning. For those with existing CPU/CNN models, MetaTF can adapt these for Akida use.
Space Cadet
Regular
And yet some are calling for a spill. When companies are looking to sign a contract, they will look at a range of things before signing. At the top of the list is the balance sheet (Does the company have the funds to be a going concern and meet their obligations which obviously includes supply of chips if that is what they are after.) The credit raise in my opinion was required to show that they could supply the chips and may very well have been a condition. Also they will check the stability of the company and board. I would think a blind man could see the damage a spill could do, especially when I feel we are getting close. By close, I personally mean next 12 months or so.
SC
Stable Genius
Regular
I invested in Beneficial Ai and now wonder if this will speed up my roi?
Dr E Brown
Regular
I think the share price fairly represents where we are in terms of revenue. If I were completely honest given the dilution with the CR and LDA it may even be construed as a little high.
Now the other part of your question - do I like it? Absolutely not, but I have clearly expressed where I hope it will be in the next two years, free from emotion, which I am in the lucky situation to be in. I understand many, perhaps most are not at the buy-in levels that I have. If it does drop further, which I sincerely hope it does not, I will buy more, based upon my earlier analysis.
I guess the bit that everybody focuses on is - does the share price suggest we should call for the removal of the CEO and the board. In my opinion, absolutely not. They cannot control the share price directly. What matters now is revenue, and relatively substantial revenue, in the 10's of millions to start with. That will move the SP. As smoothsailing stated a consumer customer or a substantial IP or chip contract will give us a significant re-rate. That is what the CEO, Board etc are and need to be focussing on, as well as maintaining our lead and opening up new markets through technological advancement. It's really a fairly simple equation.
Thanks for your question - loaded though it may have been.
Sorry if already posted. See link for comments.
www.linkedin.com
GR801 | Daniel Azevedo Novais
Frontgrade Gaisler's unveiling of the GR801 microprocessor, the inaugural offering in their GRAIN roadmap, marks a pivotal inflection point. This radiation-hardened system-on-chip ingeniously fuses the neuromorphic prowess of BrainChip's Akida technology with the stalwart NOEL-V RISC-V...
White Horse
Regular
Hi All,
This is an update on the Bascom Hunter 3U VPX module. This is in regard to new applied technology from Parallax Advanced Research led by Steven Harbour.
Titled:- From RF to Light to Spikes: A Neuromorphic Pipeline for Real-Time RF Signal Classification.
https://brainchip.com/wp-content/uploads/2026/01/From-RF-to-Light-to-Spikes_-A-Neuromorphic_IEEE-Explore.pdf
Sorry no text. This seems to be a protected PDF. The link should get you there.
This is really interesting stuff.
This is an update on the Bascom Hunter 3U VPX module. This is in regard to new applied technology from Parallax Advanced Research led by Steven Harbour.
Titled:- From RF to Light to Spikes: A Neuromorphic Pipeline for Real-Time RF Signal Classification.
https://brainchip.com/wp-content/uploads/2026/01/From-RF-to-Light-to-Spikes_-A-Neuromorphic_IEEE-Explore.pdf
Sorry no text. This seems to be a protected PDF. The link should get you there.
This is really interesting stuff.
Fullmoonfever
Top 20
Nice find thanks @White Horse
Found it interesting that BH now specify 5 x "AKD-series" not AKD1000 which from memory everything else we found / read spec'd the 1000...or am I mixed up?
Does it imply maybe moving up or has ability to move up to the 1500 / 2.0?
Also liked the key direction and guidance by a member of the USAF for stated reasons.
One of the authors in your link is from Tempo Sense.
They are quite elusive in what they are doing but quote: Our patented sensing technology blends edge processing with neuromorphic sensing, transcending the limits of traditional event sensors and imaging solutions. We’re leading a generational leap forward in the world of smart sensing.
Very little info on their webpage:
Tempo Sense - Tempo Sense
TEMPO SENSE revolutionizes Edge Computing and Sensing Technology. We’re leading a generational leap forward in the world of smart sensing.
tempo-sense.com
They released a paper called:
Harnessing the Power of AI for Smart Sensing.
I have no idea if they use AKD or not, but they were at least involved in the paper White Horse supplied above, hmm!
Hold your Horses Guzzi!One of the authors in your link is from Tempo Sense.
They are quite elusive in what they are doing but quote: Our patented sensing technology blends edge processing with neuromorphic sensing, transcending the limits of traditional event sensors and imaging solutions. We’re leading a generational leap forward in the world of smart sensing.
Very little info on their webpage:
Tempo Sense - Tempo Sense
TEMPO SENSE revolutionizes Edge Computing and Sensing Technology. We’re leading a generational leap forward in the world of smart sensing.
They released a paper called:
Harnessing the Power of AI for Smart Sensing.
I have no idea if they use AKD or not, but they were at least involved in the paper White Horse supplied above, hmm!
The author of the TS white paper is Davide Migliori, who is also an author of the White Hores/IEEE paper, and an inventor of this patent:
US12131544B2 Method for capturing motion of an object and a motion capture system 20190516
The owner is Prophesee!!!!
TS Scientific advisors Ryad Benosman and another employee were founders of Prophesee, and GrAI Matter.
Tempo sense was founded in 2024, 5 years after the patent was filed.
https://tub-rip.github.io/eventvision2025/slides/2025CVPRW_Tempo_Sense.pdf
Some interesting slides:
Akida gets a few guernseys in the IEEE paper.
Jacob Kiraz
Regular
Who is Dr Brown,
Could it be a insider
Could it be a insider
Taproot
Regular
Nice
White Horse
Regular
Hi Qman, beat me to it.
Here's the followup.
Oliver F. • 3rd+
IT Systembetreuer bei Kneer Südfenster3h
Thanks for the detailed report. What is your objective assessment of Akida? Does such a chip have a chance in the future? Will IBM use Akida? Basically, this can be the solution to save a lot of energy, right?!
· 2 replies 2 Replies on Oliver F.’s comment
Kevin D. Johnson Author
Field CTO – HPC, AI, LLM & Quantum Computing | Principal HPC Cloud Technical Specialist at IBM | Symphony • GPFS • LSF2h
Oliver F. The demo results are pretty clear. The Akida chip classified market ticks in 622 microseconds at roughly 30 milliwatts. For continuous monitoring workloads, where you are classifying millions of events and 99%+ require no action, that power profile is significant as I've laid out here and in other demos. A GPU or even a CPU performing the same classification would consume orders of magnitude more energy for an equivalent result.
I personally believe neuromorphic computing has a real future. The physics argument for event-driven spike-based processing is well established. The more precise question is which workloads justify it and how the ecosystem matures. What I can say is that for this specific pattern, continuous classification as a gate for downstream compute, it proved to be the right tool. Neuromorphic compute like Akida saves on scarce resources like power while extending what can be done in line with Symphony's capabilities.
On the IBM question, I can't speak to IBM's product plans. The market typically drives that conversation.
“Neuromorphic compute like Akida saves on scarce resources like power while extending what can be done in line with Symphony's capabilities.
On the IBM question, I can't speak to IBM's product plans. The market typically drives that conversation.”
Why can’t people just commit and call it what it is?
If he says “like Akida,” but the thing he’s clearly convinced by (and the results he keeps pointing to) are Akida, then… why not say Akida?
Especially because, realistically, there isn’t a true “Akida-equivalent” out there in the same sense. So when someone writes “like Akida,” it raises a fair question:
What other systems does he actually mean?
What has he achieved similar results with, if not Akida?
Is this just hedging language, and if yes…why?
No hate at all …. I’m genuinely trying to understand the wording. If the claim is basically “right now only Akida can do this (at least in his experience),” then saying “like Akida” sounds unnecessarily vague.
So yeah: I’m just asking … what are people afraid of when they use “like Akida” instead of simply saying “Akida”?
On the IBM question, I can't speak to IBM's product plans. The market typically drives that conversation.”
Why can’t people just commit and call it what it is?
If he says “like Akida,” but the thing he’s clearly convinced by (and the results he keeps pointing to) are Akida, then… why not say Akida?
Especially because, realistically, there isn’t a true “Akida-equivalent” out there in the same sense. So when someone writes “like Akida,” it raises a fair question:
What other systems does he actually mean?
What has he achieved similar results with, if not Akida?
Is this just hedging language, and if yes…why?
No hate at all …. I’m genuinely trying to understand the wording. If the claim is basically “right now only Akida can do this (at least in his experience),” then saying “like Akida” sounds unnecessarily vague.
So yeah: I’m just asking … what are people afraid of when they use “like Akida” instead of simply saying “Akida”?
Similar threads
