BRN Discussion Ongoing

[Food4 thought]

I asked Gemini While BrainChip is actively engaging with numerous companies and has established several partnerships, the specific details of which companies are currently shipping commercial products incorporating BrainChip's technology can be proprietary and are not always publicly disclosed.
However, based on recent announcements and partnerships, we can identify several companies and areas where BrainChip's Akida technology is being actively integrated or evaluated:
Confirmed Partnerships & Collaborations:
* RTX's Raytheon: Partnering to service a contract with the Air Force Research Labs for radar signal processing using neuromorphic chips.
* VVDN Technologies: Collaborated on the BrainChip Akida Edge AI Box and can develop custom versions for OEMs.
* Vedya Labs: Developing video object detection models for the Akida Edge AI Box.
* Intel Foundry Services (IFS): BrainChip is an IP partner, making their technology accessible to chip designers using Intel's manufacturing services.
* Prophesee: Collaborating to optimize computer vision AI performance by integrating Akida with Prophesee's Metavision platform.
* MegaChips: Licensing Akida IP to incorporate into their ASIC solutions for various markets like automotive, IoT, and industrial robotics.
* MYWAI: Partnering to deliver Edge AI solutions for industrial and robotic sectors, integrating Akida with MYWAI's AIoT platform.
* Teksun: Integrating Akida into their product development channels.
* Unigen: Providing end-to-end solutions for the Cupcake Edge AI Server, powered by BrainChip.
* GMAC Intelligence: Validating Akida for facial, vehicular, and voice recognition in quick-service restaurant applications.
* Tata Elxsi: Integrating Akida technology into medical devices and industrial applications.
* Eastronics & SalesLink: Sales partners expanding the reach of Akida technology in Europe and Israel across various applications.
* AI Labs Inc.: Collaborating on next-generation application development leveraging BrainChip's Minsky AI technology.
* Frontgrade Gaisler & Airbus Defence and Space: Partnerships focused on leveraging Akida for ultra-low-power AI in space applications under a European Space Agency initiative.
* NVISO: Partnering on human behavioral analytics for automotive and edge AI devices.
Target Markets and Potential Integrations:
BrainChip's technology is being targeted for a wide range of applications across various industries, suggesting potential integrations with numerous companies in these sectors:
* Automotive: Enhancing safety and in-cabin experiences with on-chip learning for features like driver monitoring and autonomous driving.
* Industrial IoT: Improving predictive maintenance, anomaly detection, and automation in manufacturing and other industrial settings.
* Home & Consumer: Enabling intelligent features in devices like smart thermostats, security systems, and entertainment systems through efficient processing of various sensor modalities (vision, audio, etc.).
* Health & Wellness: Developing cost-effective portable diagnostic devices and enabling real-time vital signs prediction.
* Security & Surveillance: Enhancing object detection, facial recognition, and anomaly detection in security systems with low-power edge processing.
* Aerospace: Developing ultra-low-power AI for space missions and other aerospace applications.
* Smart Cities: Enabling efficient management of resources and services through intelligent sensors and edge processing.
* Agriculture: Optimizing resource use through precision agriculture techniques powered by edge AI.
* Robotics: Enhancing the intelligence and autonomy of robots through efficient on-device AI processing.
Key Takeaway:
While a definitive list of all current commercial deployments is not fully public, BrainChip has established significant partnerships and is actively working with companies across various sectors to integrate its Akida neuromorphic technology into next-generation products and solutions, particularly those requiring low-power, high-performance edge AI capabilities. The listed partnerships provide a strong indication of the companies currently at the forefront of adopting BrainChip's innovations.

 

[manny100]

Discussion considerations prior to AGM voting:
Will AI at the Edge grow exponentially? Industry experts say it will.
If you disagree with the experts, please provide reasons.
Will BrainChip play a good sized part in the Edge AI expected exponential growth?
The general consensus is that the Edge will play an enormous part in daily lives in due course.
There will be chip choices but AKIDA/TENNs and Tech interest since Sept'24 by clients suggest we will do well.
If you do not believe BRN will not play a decent sized role in Edge AI growth, please explain your reasons. If you have suggestions of Edge companies providing event-based learning on chip with extreme low power, is commercially available and has proven superiority to AKIDA/TENNs/Pico it would be great if you could let us know.
If you believe what the experts are saying about Edge AI exponential growth then our growing patent portfolio would have a value many times the current SP.
Pitt Street valued it at $1.59 in June'24 based on the Intel acquisition Habana and we are currently way more advanced than they were at the time of acquisition. Yes , BRN paid for the report but the $1.59 was based on an actual sale (Habana to Intel) and is an accepted valuation method.
BrainChip_intiation_report_2024_25_6_2024.pdf
Given the flurry of client and tech positive news since Sep't'24 we probably will see an updated report soon.

 

[Diogenese]

Hallucinogenic software solutions for mission critical hardware sounds like the go There’s a thing called TENNs which doesn’t hallucinate! I fail to see how ChatGPT could ever be applied in any important way to mission critical systems - it would fail the extensive, stringent testing that such systems require.

Hi jt,

The reference is to OpenAI's "advanced models", which may not have the same faults as ChatGPT.

 

[7für7]

I’ve been invested in BrainChip since the early days… way before AKIDA even has been commercialised. As a sci-fi fan and someone who genuinely believes in the positive potential of AI, I see huge opportunities in Edge AI, and especially in AKIDA, to truly improve people’s lives.

That said, as an investor, I’m honestly a bit puzzled. We’ve seen steady technical progress, a growing patent portfolio, industry recognition… and yet the share price keeps sliding. I get that the company is focused on aerospace and defense…. cool tech, sure, but from an investor’s perspective, what’s missing is the push toward real world, massmarket applications that make a tangible difference in daily life.

Personally, I’d love to see BrainChip move more into Wearables that support the hearing- or visually-impaired, AI systems that help older people stay safe and independent, or just smart devices that make everyday life more convenient for all of us!

These are the kinds of things people can actually feel. And they’d show the market that AKIDA isn’t just technically impressive, but also socially relevant and commercially scalable.

I get that frustration is high right now. But I still believe, if BrainChip can land just one or two serious licensing deals in the consumer or assistive tech space, the tide could turn. That’s when the market will wake up and potential becomes reality.

Mass production is the key!

 

[7für7]

Wow impressive solutions

Edge Computing Solutions and Technologies - Intel

Learn how Intel edge computing solutions and technologies streamline integration and get your network deployed faster.

www.intel.com

 

[7für7]

Wow impressive solutions

Edge Computing Solutions and Technologies - Intel

Learn how Intel edge computing solutions and technologies streamline integration and get your network deployed faster.

www.intel.com

Even HP is addressing the topic of edge computing …a clear sign that a significant market shift is approaching. Companies are already preparing their customers for the coming change.

https://www.hpe.com/emea_europe/en/what-is/edge-computing.html

 

[manny100]

I’ve been invested in BrainChip since the early days… way before AKIDA even has been commercialised. As a sci-fi fan and someone who genuinely believes in the positive potential of AI, I see huge opportunities in Edge AI, and especially in AKIDA, to truly improve people’s lives.

That said, as an investor, I’m honestly a bit puzzled. We’ve seen steady technical progress, a growing patent portfolio, industry recognition… and yet the share price keeps sliding. I get that the company is focused on aerospace and defense…. cool tech, sure, but from an investor’s perspective, what’s missing is the push toward real world, massmarket applications that make a tangible difference in daily life.

Personally, I’d love to see BrainChip move more into Wearables that support the hearing- or visually-impaired, AI systems that help older people stay safe and independent, or just smart devices that make everyday life more convenient for all of us!

These are the kinds of things people can actually feel. And they’d show the market that AKIDA isn’t just technically impressive, but also socially relevant and commercially scalable.

I get that frustration is high right now. But I still believe, if BrainChip can land just one or two serious licensing deals in the consumer or assistive tech space, the tide could turn. That’s when the market will wake up and potential becomes reality.

Mass production is the key!

I think it's only a matter of time before someone comes up with AKIDA/TENNs based effective visual aid. Perhaps aided by Prophesee?
The great thing about AKIDA is if you buy it you know BRN will keep improving/updating it.

 

[DingoBorat]

well one thing that the result of less thumbs up on FF on crapper is that there are 1000 eys on TSE and many of these eyes don't waste time on crapper anymore so that's also a plausible reason for less thumbs up to FF. So i prob would add a few hundred extra thumbs up to FF but yes i guess that still slightly down as now we probably want a rocket instead of what's a questionable finger.

No.
The "1000" eyes (which is "only" 500 people, but they are "actively" scouring the Net, for info and links on BRN) may have been "close" to true at some time, 2 to 3 years ago (probably closer to 2000 eyes, with a keen multi-daily interest in all things BrainChip).

One of the reasons FactFinder left the forum (other than him being dakked by Frangipani, and him having nothing to say about his shivelling integrity, in the way he was increasingly expressing himself).
Which was obviously the major factor..
Was a post I did in conjunction with figures from ZeeB0t, on the amount of daily traffic, on this site, which is definitely not 500 people, it's much much less.

You can basically gauge it, from the number of "likes" a "breakout" post gets here (combined with how quickly it reaches that number).

FactFinder, thought he had the "eyes of the World" on this forum, in ever increasing numbers, due to various progress of the Company.

The fact is, when he decided to take his over a year (or whatever it was) sabbatical (which happened, for whatever reason, to directly coincide with BRN's inclusion and kicking out, from the ASX200) he lost much of his "flock".
And the audience, simply wasn't here anymore, for the prodigious amount of brilliant effort, he was putting into numerous daily posts.

Since he's been back on the crapper (and literally forced to clean up his act, because they are only too willing to call things out there) he and the Company, have not been able to draw back the retail investor interest.

BrainChip, is just another "hopeful speculative company" there.

Yes, there are probably still close to 40000 shareholders (which is Huge) but most holdings are most likely under $500 worth (an unnecessary burden on the Company) and have been well and truly "bottom drawed".


I'm not having a go at Frangipani, as she was "right" in what she pointed out, even if most disagree with her "attacking" or "questioning" who had become almost a BRN forum "God", nor am I attacking FactFinder, who I know, was/is of huge value to BrainChip's investor sentiment and undeniably, the best poster, this stock has had.


It's a hard task though, for any "prophet" or "cult leader" (to use the most positive and negative connotations) to keep his "followers" hanging on, when the Company continues to "fail" in any real commercial progress.

I do believe interest will "Fire Up" dramatically, as I've said and within a relatively short period, if BrainChip "does" have a major commercial breakthrough, but it won't be a dramatic flare up like before, but like burning damp wood..

It will get there though.

 

[DingoBorat]

It appears not everyone at AFRL resp. collaborating with AFRL would agree that Akida is always the superior choice:

Anomaly detection of CAN bus messages on neuromorphic hardware | SPIE Optics + Photonics

View presentations details for Anomaly detection of CAN bus messages on neuromorphic hardware at SPIE Optics + Photonics

spie.org

View attachment 82236


View attachment 82235
View attachment 82234

Tarek Taha

View attachment 82233

I reckon that guy's full of shit Frangipani..
I don't care how many letters he has after his name.

I don't think there's any way, that AKIDA properly configured, would be 91 times less efficient than LoiHi.

Either he's doing something very wrong, or his research is being "paid" for by Intel.

 

[manny100]

No.
The "1000" eyes (which is "only" 500 people, but they are "actively" scouring the Net, for info and links on BRN) may have been "close" to true at some time, 2 to 3 years ago (probably closer to 2000 eyes, with a keen multi-daily interest in all things BrainChip).

One of the reasons FactFinder left the forum (other than him being dakked by Frangipani, and him having nothing to say about his shivelling integrity, in the way he was increasingly expressing himself).
Which was obviously the major factor..
Was a post I did in conjunction with figures from ZeeB0t, on the amount of daily traffic, on this site, which is definitely not 500 people, it's much much less.

You can basically gauge it, from the number of "likes" a "breakout" post gets here (combined with how quickly it reaches that number).

FactFinder, thought he had the "eyes of the World" on this forum, in ever increasing numbers, due to various progress of the Company.

The fact is, when he decided to take his over a year (or whatever it was) sabbatical (which happened, for whatever reason, to directly coincide with BRN's inclusion and kicking out, from the ASX200) he lost much of his "flock".
And the audience, simply wasn't here anymore, for the prodigious amount of brilliant effort, he was putting into numerous daily posts.

Since he's been back on the crapper (and literally forced to clean up his act, because they are only too willing to call things out there) he and the Company, have not been able to draw back the retail investor interest.

BrainChip, is just another "hopeful speculative company" there.

Yes, there are probably still close to 40000 shareholders (which is Huge) but most holdings are most likely under $500 worth (an unnecessary burden on the Company) and have been well and truly "bottom drawed".


I'm not having a go at Frangipani, as she was "right" in what she pointed out, even if most disagree with her "attacking" or "questioning" who had become almost a BRN forum "God", nor am I attacking FactFinder, who I know, was/is of huge value to BrainChip's investor sentiment and undeniably, the best poster, this stock has had.


It's a hard task though, for any "prophet" or "cult leader" (to use the most positive and negative connotations) to keep his "followers" hanging on, when the Company continues to "fail" in any real commercial progress.

I do believe interest will "Fire Up" dramatically, as I've said and within a relatively short period, if BrainChip "does" have a major commercial breakthrough, but it won't be a dramatic flare up like before, but like burning damp wood..

It will get there though.

Over on the crapper FF posts are generally links to positive articles of interest for holders with some comment rarely being an outright 'buy comment'.
Anything positive on the crapper is 'open season for the pistol' club. That is expected and ok as all posters have their own agendas' to push.
You just have to sort through them all.
In the end market participants just have to decide:
Is AI at the Edge going to have exponential growth as tipped by industry experts, and
If so, will BRN will have a meaningful role to play in that growth.
If the answer to both is yes then a 'patient' punt at these prices may be considered.
If no to both then avoid.
There is plenty of research material available.

 

[jtardif999]

Hi jt,

The reference is to OpenAI's "advanced models", which may not have the same faults as ChatGPT.

Yes @Diogenese but OpenAI’s tech would also still potentially hallucinate since ChatGTP models are not created utilising orthogonal polynomials in the way TENNs is. In the recent podcast with BRN technical leads explained how orthogonal polynomials create models that can only utilise real world objects making the models reliable and resilient against the potential for hallucination which is inherent in models created using traditional transformer methods - such as the models used by ChatGPT. Now, having said that I am assuming that OpenAI would still adopt a similar, baselined approach to develop more advanced models that would have the potential to hallucinate since it would not be an entirely new approach to create them.

 

[Labsy]

So irrelevant that I correctly predicted the last two massive STRIKE votes against management when every one here said it would,nt happen, Like management you can’t read the room mate, Shareholders say NO and will absolutely say no to moving to the USA.

Don't think so...I'll be voting in the affirmative for everything. Sorry mate, your and your buddies vote may sting a little...that's all. Happy to buy your shares off market for 20c if you want to dump them...just offering

 

[MDhere]

Having watched it also, the presenter merely acknowledges Brainchip, along with other companies, offers a neuromorphic hardware systems. I think the basis of the presentation, she and her students are trying to create a more efficient neuromorphic system for commercial application. She is basically saying current systems are too clunky to work with and are not efficient or reliable. Go figure. I am not sure who is telling the truth. The professor presenter or Brainchip sales department. It’s a tough one to watch as she is saying current hardware isn’t cutting the mustard. Yet, we have many engagements that may say otherwise. Time will tell I guess. Bit of an embellishment drawing a conclusion from the presentation that Brainchip is capital of neuromorphic.

I agree UA , i think the presenter is pushing for the "Thor: adoption like a similar set up to edge impulse. But to name Brainchip at the start I thought at least as a positive. In regards to embellished drawing the conclusion as us as the capital of neuromorphic might be embellished but its not completely unrealistic, well not in my mind it isn't

 

[MDhere]

No.
The "1000" eyes (which is "only" 500 people, but they are "actively" scouring the Net, for info and links on BRN) may have been "close" to true at some time, 2 to 3 years ago (probably closer to 2000 eyes, with a keen multi-daily interest in all things BrainChip).

One of the reasons FactFinder left the forum (other than him being dakked by Frangipani, and him having nothing to say about his shivelling integrity, in the way he was increasingly expressing himself).
Which was obviously the major factor..
Was a post I did in conjunction with figures from ZeeB0t, on the amount of daily traffic, on this site, which is definitely not 500 people, it's much much less.

You can basically gauge it, from the number of "likes" a "breakout" post gets here (combined with how quickly it reaches that number).

FactFinder, thought he had the "eyes of the World" on this forum, in ever increasing numbers, due to various progress of the Company.

The fact is, when he decided to take his over a year (or whatever it was) sabbatical (which happened, for whatever reason, to directly coincide with BRN's inclusion and kicking out, from the ASX200) he lost much of his "flock".
And the audience, simply wasn't here anymore, for the prodigious amount of brilliant effort, he was putting into numerous daily posts.

Since he's been back on the crapper (and literally forced to clean up his act, because they are only too willing to call things out there) he and the Company, have not been able to draw back the retail investor interest.

BrainChip, is just another "hopeful speculative company" there.

Yes, there are probably still close to 40000 shareholders (which is Huge) but most holdings are most likely under $500 worth (an unnecessary burden on the Company) and have been well and truly "bottom drawed".


I'm not having a go at Frangipani, as she was "right" in what she pointed out, even if most disagree with her "attacking" or "questioning" who had become almost a BRN forum "God", nor am I attacking FactFinder, who I know, was/is of huge value to BrainChip's investor sentiment and undeniably, the best poster, this stock has had.


It's a hard task though, for any "prophet" or "cult leader" (to use the most positive and negative connotations) to keep his "followers" hanging on, when the Company continues to "fail" in any real commercial progress.

I do believe interest will "Fire Up" dramatically, as I've said and within a relatively short period, if BrainChip "does" have a major commercial breakthrough, but it won't be a dramatic flare up like before, but like burning damp wood..

It will get there though.

Roger that, Dingo

 

[manny100]

ANT61 Inks Commercial Agreement with Japan's Sompo Risk Management

Australian satellite hardware company ANY61 has signed its first commercial agreement with Japan’s Sompo Risk Management. The agreement marks a significant milestone for ANT61 as it expands to the Japanese market.

spaceanddefense.io

I see ANT61 has made sales. See link.
" Australian satellite hardware company ANT61 has signed its first commercial agreement with Japan’s Sompo Risk Management. The agreement marks a significant milestone for ANT61 as it expands to the Japanese market."
" The first generation of the ANT61 Brain uses the BrainChip AkidaTM chip for power-efficient AI and is currently on board the Optimus-1 satellite, which was deployed recently by the SpaceX Transporter-10 mission. We will test ANT61 Brain later this year and perform training and evaluation of various neural networks that we will use for future in-orbit servicing technology demonstrations."

Revolutionizing Space Infrastructure: ANT61 Smart Robots

Revolutionizing Space Infrastructure: ANT61 Smart Robots April 8, 2024 – Mikhail Asavkin Space infrastructure plays a key role in maintaining civilization on Earth. From moving goods around the globe and monitoring the effects of global warming, such as forest fires and floods, to everyday...

brainchip.com

 

[Diogenese]

Yes @Diogenese but OpenAI’s tech would also still potentially hallucinate since ChatGTP models are not created utilising orthogonal polynomials in the way TENNs is. In the recent podcast with BRN technical leads explained how orthogonal polynomials create models that can only utilise real world objects making the models reliable and resilient against the potential for hallucination which is inherent in models created using traditional transformer methods - such as the models used by ChatGPT. Now, having said that I am assuming that OpenAI would still adopt a similar, baselined approach to develop more advanced models that would have the potential to hallucinate since it would not be an entirely new approach to create them.

Yes. Hallucinations are a problem.

Just using video* as an example, I wonder if the long skip introduced in Akida 2 could be a remedy for some hallucinations. My thinking is that, once a segment of an image is classified in one layer, then the block of neurons corresponding to the identified image segment are passed to the output layer, bypassing further processing by subsequent layers. This is just my guess at how long skip works, based on the assumption that hallucinations are caused, at least in part, by overfitting already classified blocks of neuron in subsequent layers.

Note this is just my guess as to the relation between long skip, overfitting and hallucination. I don't have any citations to support this ... yet.

I don't know what OpenAI's "advanced models" involve, but it would be advisable for them to be looking at alternative techniques, so I would not write off the possibility that OpenAI are developing such alternatives.

PS: So it seems that overfitting is one cause of hallucinations:

https://www.datacamp.com/blog/ai-hallucination

Causes of AI Hallucinations​

Four key factors often contribute to hallucinations:

1. Insufficient or biased training data
2. Overfitting
3. Faulty model architecture
4. Generation methods

* I use video as an example because I find it easier to grasp (eg the multi-labrador/retriever image). NLP is even further beyond my comprehension.

PPS: I think another possible contributor to hallucination could be error backpropagation**. Akida uses feedforward ML.

** https://www.intel.com/content/www/u...rphic-computing-loihi-2-technology-brief.html

Enhanced learning capabilities.
Loihi primarily supported two-factor learning rules on its synapses, with a third modulatory term available from nonlocalized “reward” broadcasts. Loihi 2 allows networks to map localized “third factors” to specific synapses. This provides support for many of the latest neuroinspired learning algorithms under study, including approximations of the error backpropagation algorithm, the workhorse of deep learning. While Loihi was able to prototype some of these algorithms in proof-of-concept demonstrations, Loihi 2 will be able to scale these examples up, for example, so new gestures can be learned faster with a greater range of presented hand motions.


PPPS:
Here's a train hallucination:

https://en.wikipedia.org/wiki/Hallucination_(artificial_intelligence)

As you say, TENNs, with it's temporal analytics, should be able to suppress such hallucinations, as this double image is clearly derived from different points in time.

 

[Pom down under]

Apart from being a great song.

 

[rgupta]

Rgupta that has always and your comments are quiet valid been a bugbear of mine two, I had assumed that they thought they must have been close to contracts and did,nt feel the need to raise a large amount, but unfortunately management always stated that they had enough for a two year runway and they did,nt feel the need to raise further funds,but strangely the money disapaited that fast it was,nt funny,shareholders were caught out by capital raises that they said were,nt needed. But the shorters knew they were coming did,nt they! Coincidence I think not.

Yes it may be a coincidence as well. Just look back what happened at that time and how sp of above 2 was again a coincidence. Antonio told us akida 1000 was not good enough to support the sp.
Sean told us he delivered a five year plan,
Shorters lift the share price from $1.2 to 2.34 and then started shorting there. When everyone thinks 50 cents is the bottom shorters raised their game and end it @15 cents.
Management should understand the complications of ASX 200 company. But yes everything is coincidental. If there are so many coincidences there to happen then what is the role of a manager.
I don't mind all the games played as on date but where is the light on other end of the tunnel??
Sorry to me it is a below average game of a manager. If he was so confident and now his confidence is shattered that means he should tell the board and holders about the same. But the board is advocating 7.5 million RSU which is 120% of achievement for that big coincidence.
We holders have reasons to ask questions.
Dyor

 

[jrp173]

Yes it may be a coincidence as well. Just look back what happened at that time and how sp of above 2 was again a coincidence. Antonio told us akida 1000 was not good enough to support the sp.
Sean told us he delivered a five year plan,
Shorters lift the share price from $1.2 to 2.34 and then started shorting there. When everyone thinks 50 cents is the bottom shorters raised their game and end it @15 cents.
Management should understand the complications of ASX 200 company. But yes everything is coincidental. If there are so many coincidences there to happen then what is the role of a manager.
I don't mind all the games played as on date but where is the light on other end of the tunnel??
Sorry to me it is a below average game of a manager. If he was so confident and now his confidence is shattered that means he should tell the board and holders about the same. But the board is advocating 7.5 million RSU which is 120% of achievement for that big coincidence.
We holders have reasons to ask questions.
Dyor

Regarding your comment of "management should understand the complications of ASX 200 company", I agree 100% with you.

We have been on the ASX since 2016, for management to now be trying to say that that ASX does not work for them is total nonsense and insulting to Australian shareholders.

We've been on the ASX for 9 years, and NOW they decide it's not for them. If they truly believe this, why did they not decide to do a move when our share price was higher, so that shareholders would not be subject to such a huge consolidation.

Management have a lot to answer for in my opinion.

 

[JB49]

Listen from 35min. Yann LeCun briefly talks about analog vs digital spiking neural networks. He then refers to some of his colleagues at Meta looking into SNNs for the glasses due to power consumption problems.

There have been a few job adverts with meta referring to SNNs recently.