BRN Discussion Ongoing

[manny100]

https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-433475

https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-433479

Hi @DingoBorat and @Diogenese

AGM May 2024 -

Tony Lewis: well so TENNs is a competitor against, existing neural networks, our customers have a view of the world that is pre TENNs they have to be introduced to the technology and they're being introduced in a stage manner we're going out we're talking about TENNs we're publishing papers we're getting people familiarised I believe that once they're familiarised with TENNs they're going to see that we can use it for all sorts of things so we're proving out how we can use it for audio de-noising as you saw there but also ASR is on our agenda, processing biomedical samples that's on our agenda and also these large language models and so we're proving out what we can do with TENNs and once we have those proof points we can go to customers and say hey guys this is going to be you should go with TENNs because it's going to be a lot more power efficient than what you're using right now it's going to be good for you eat your vegetables and you know learn about our technology.

Sean: before the next question comes a couple of themes that came out if you heard my speech, we were going to put more resources into TENNs so more development.

Tony’s comments from just three months ago, along with the recent CR, SPP, and investor presentation from two weeks ago, make it hard for me to believe that we’ll be waking up to a trading halt involving the sale of TENNs IP or software anytime soon.

I like the team's efforts in building TENN's profile through articles, white paper, etc and by ensuring that there are multiple demos, models and possibly even simulation environments are ready before moving towards commercialisation. These lessons have clearly been learned from the experience with AKIDA when introducing a brand-new technology.

Late in the Quarterly update with Tony, Sean spoke about TENNS improvements and said support for LLM's at the Edge would be available very soon. That could be IMO potentially huge in that it can take more work away from the cloud and position it to non cloud.
How soon is very soon? That does not mean sales soon but it provides another marketing tool. Regardless of the LLM plans i expect we will see some developments via the R & D cash.
Also I expect that Gen 3 should become a reality in 2025 hopefully in the 1st half half.

 

[Flenton]

Our chips are currently being used for electric vehicles, home security and surveillance, home intelligence and maintenance, manufacturing maintenance and management, robotics, automation, quality control, healthcare data, and fitness equipment!

Where and how are they being used? Must not be in a commercial product or we'd be seeing revenue.

 

[DingoBorat]

Wow, the development is fast ...

x.com

x.com

Wait until you see Optimus 3, probably near the end of this year..

 

[CHIPS]

Wait until you see Optimus 3, probably near the end of this year..

Yes, I can't wait to see how it will develop. I am also invested in Tesla because of Optimus.

 

[Diogenese]

Wait until you see Optimus 3, probably near the end of this year..

That's a whole bunch of accelerometers and gyroscopes that could use an SNN.

 

[CHIPS]

WOW ... here it is described. I heard LLM, screen in the face, doubled number of utilities, larger battery pack size, etc.
There MUST BE a need for Akida and TENNs soon.

 

[Terroni2105]

I know they've been advertising but looks like maybe using recruitment firm now too (not searched if cybercoders posted prev) and still trying or recently have fill the role.

Not my bold / highlighting.

Wonder how much poetic licence has been taken re where the chips are "currently" being used or if that is a Freudian slipped through

LinkedIn post HERE

Website

machine learning engineer Jobs | CyberCoders

Work with top companies as a machine learning engineer. View and apply to these listings or browse for similar jobs.

www.cybercoders.com

Machine Learning Engineer - 130K-170K - Hybrid in Laguna Hills​

CyberCoders Mission Viejo, CA​

2 weeks ago 136 applicants​

Job Title: Machine Learning Engineer

Location: Hybrid in Laguna Hills, CA - 3 days onsite

Salary: $130,000 - $170,000+ Annual Base Salary (DOE), Large RSU's Package, Full Benefits, PTO, Bonuses, 401K

Requirements: 3+ years of Machine Learning experience, experience training large models with large numbers of GPUs, solid math foundation to comprehend what research scientists are creating

We are a well funded, small company that has been around for over 10 years creating chips that hold AI technology. Our chips are currently being used for electric vehicles, home security and surveillance, home intelligence and maintenance, manufacturing maintenance and management, robotics, automation, quality control, healthcare data, and fitness equipment!

Due to our increased workload, we are currently looking to add a Machine Learning Engineer to our experienced team in Laguna Hills, CA! We have about 40 employees but this person will be on our Machine Learning Team reporting to the CTO (about 5 engineers). This is an opportunity to develop AI using new algorithms that people have not seen before for new hardware that no one has programmed before! We are looking for talented candidates that have studied subjects like Theoretical Physics or AI.

Responsibilities

• Implement and optimize ML Algorithms for deployment on embedded systems
• Collaborate with research team to translate ML models
• Develop and maintain code in python, C, and C++ for real time systems
• Stay up to date with ML, embedded AI, and more AI/ML related technologies
• Enhance system performance by working on ML algorithm and hardware co-design
• Debug software to make sure it is performing efficiently on Akida
• Customer interfacing and intake to understand needs and problems, provide tech support for ML applications
• Help develop Akida software

Qualifications

• 3+ years of professional experience in ML
• Python
• C++
• RTOS
• Embedded AI
• Bachelor's Degree in Computer Science or Electrical Engineering

No free lunch ?

 

[DingoBorat]

That's a whole bunch of accelerometers and gyroscopes that could use an SNN.

Neuromorphic computing is an absolute perfect fit, for humanoid robots.

These things run on batteries obviously and you can't just chuck a big one in, like you can an EV..

It's a virgin market, with incredibly huge upscale, I can see other car companies, gravitating towards making humanoid robots as well.
Hyundai, is obviously heading that way.
Honda, has dabbled in the past.
A Mercedes humanoid robot, would be a stand out.

Just look at how complicated cars are, whether ICE or EV, with all the different systems and supply chains involved, compared to a humanoid robot, once developed.

The profit margins will likely be higher and they would be able to pump them out at a faster rate.

I really don't understand, how car companies survive and what type of market there is, for people buying "brand new" cars..
And that's actually evidenced, by the sheer number of car companies, that have either failed, or been saved, basically just for the marketing strength of their heritage.

Whack a Lamborghini badge, on a VW..

Neuromorphic has a guaranteed Future, in humanoid robotics and large scale production, isn't as far off, as some may think..

 

[DingoBorat]

https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-433475

https://thestockexchange.com.au/threads/brn-discussion-ongoing.1/post-433479

Hi @DingoBorat and @Diogenese

AGM May 2024 -

Tony Lewis: well so TENNs is a competitor against, existing neural networks, our customers have a view of the world that is pre TENNs they have to be introduced to the technology and they're being introduced in a stage manner we're going out we're talking about TENNs we're publishing papers we're getting people familiarised I believe that once they're familiarised with TENNs they're going to see that we can use it for all sorts of things so we're proving out how we can use it for audio de-noising as you saw there but also ASR is on our agenda, processing biomedical samples that's on our agenda and also these large language models and so we're proving out what we can do with TENNs and once we have those proof points we can go to customers and say hey guys this is going to be you should go with TENNs because it's going to be a lot more power efficient than what you're using right now it's going to be good for you eat your vegetables and you know learn about our technology.

Sean: before the next question comes a couple of themes that came out if you heard my speech, we were going to put more resources into TENNs so more development.

Tony’s comments from just three months ago, along with the recent CR, SPP, and investor presentation from two weeks ago, make it hard for me to believe that we’ll be waking up to a trading halt involving the sale of TENNs IP or software anytime soon.

I like the team's efforts in building TENN's profile through articles, white paper, etc and by ensuring that there are multiple demos, models and possibly even simulation environments are ready before moving towards commercialisation. These lessons have clearly been learned from the experience with AKIDA when introducing a brand-new technology.

When we may secure an IP deal for them, wasn't the subject of my post.

It was the fact, that you were wrong, about AKIDA 2.0 TENNS IP, TENNS software and TENNS Pleaides software, not being part of BrainChip's current IP portfolio, correct?

Which you confidently stated as fact.

Saying you don't see an IP deal for them anytime soon, is your opinion and is neither right nor wrong and you are of course, free to have.

Saying something is not a product, when it is and wrongly stating that it's because it's still in development, is not an opinion, as you are stating it as a fact, which is not okay.

 

[Diogenese]

When we may secure an IP deal for them, wasn't the subject of my post.

It was the fact, that you were wrong, about AKIDA 2.0 TENNS IP, TENNS software and TENNS Pleaides software, not being part of BrainChip's current IP portfolio, correct?

Which you confidently stated as fact.

Saying you don't see an IP deal for them anytime soon, is your opinion and is neither right nor wrong.

Saying something is not a product, when it is and wrongly stating that it's because it's still in development, is not an opinion, as you are stating it as a fact, which is not okay.

The basic TeNNs concept was described in a patent application filed over 2 years ago. BRN would have already had simulation software before filing the patent application. It already has the basic functionality which is being demonstrated to potential customers and further capabilities will be revealed in a staged manner. By that, I understand that BRN are demonstrating the current functionality such as audio denoising, and, as new functionality such as ASR and LLMs are developed, this will be provided to the clients. That's the beauty of software. It can be continuously upgraded without needing a new chip.

The cloud-based FPGA is being developed to give a hands-on experience for customers to get a feel for using it, although they will not get the real speed experience, but these customers will understand the interweb delay.

Anil announced that Akida 2 was ready for tape-out several months ago. TeNNs is an integral part of Akida 2. Company "X", (not the one "formerly known as"), was said to have taken charge of the tape-out/manufacture. BRN would not have prepared for tape-out if there was not a sufficiently operative device.

Given tape-out readiness, the interesting possibility is that the IP for silicon may already be in a state to run LLMs etc, but the models need to be developed.

Goosebumps ...

Footnote: Again, speculation by uninformed sources should not be used as the basis for investment decisions.

 

[DingoBorat]

WOW ... here it is described. I heard LLM, screen in the face, doubled number of utilities, larger battery pack size, etc.
There MUST BE a need for Akida and TENNs soon.

I'm waiting for official news, on the Tesla channel..

This came up on my feed, after watching your clip and is a good watch..



I've been trying to keep abreast of humanoid robotic development, but completely missed the introduction of Figure 02, on the 6th of August, which is already being put to use, in BMW factories..



Opening sequence, is copying, for obvious comparison, almost the exact movements displayed by Optimus 2.

Unfortunately, it also has Optimus 2's "just shit my pants" walk..
Which Elon would be well aware of the ridicule of, as someone's comment on it, got over 46000 thumbs up (I was one ) and is the Top comment.

So I'm betting, it will be one of the improvements, in Optimus 3, I mean even Unitree G1 walks better..

 

[cosors]

Where is the little willy?

 

[rgupta]

The basic TeNNs concept was described in a patent application filed over 2 years ago. BRN would have already has simulation software before filing the patent application. It already has the basic functionality which is being demonstrated to potential customers and further capabilities will be revealed in a staged manner. By that, I understand that BRN are demonstrating the current functionality such as audio denoising, and, as new functionality such as ASR and LLMs are developed, this will be provided to the clients. That's the beauty of software. It can be continuously upgraded without needing a new chip.

The cloud-based FPGA is being developed to give a hands-on experience for customers to get a feel for using it, although they will not get the real speed experience, but these customers will understand the interweb delay.

Anil announced that Akida 2 was ready for tape-out several months ago. TeNNs is an integral part of Akida 2. Company "X", (not the one "formerly known as"), was said to have taken charge of the tape-out/manufacture. BRN would not have prepared for tape-out if there was not a sufficiently operative device.

Given tape-out readiness, the interesting possibility is that the IP for silicon may already be in a state to run LLMs etc, but the models need to be developed.

Goosebumps ...

Footnote: Again, speculation by uninformed sources should not be used as the basis for investment decisions.

I assume software support is the biggest reason for brainchip not to proceed with akida 2.
To me we are not a chip manufacturer and have a very small ecosystem to sell our chips.
But if a third party big company produce the chips then they will have good exposure as well as ecosystem to develop softwares quickly.
Akida 1000 was an eye opener for brainchip and by now they know the limitation of chip manufacturing and selling.
So I will say it is a wise decision by company not to produce any more chips but just support someone who can manufacture and sell the same.

 

[TECH]

That's a whole bunch of accelerometers and gyroscopes that could use an SNN.

The education process sometimes ends up with a class full of students all sitting at the back wearing the same pointed hat
with a big D on it....currently the handbrake is still on when it comes to Peters et al futurist architecture.....they (market leaders)
are either changing very slowly, resisting the change that will be forced upon them, as it is only going to take ONE MAJOR
PLAYER to actually bring a successful product to market and the rest of the flock will HAVE to follow or die a slow death with the Von.

This is why the University program we have established is the gateway to the future, very smart play by Peter et al, also you may have
seen how many applicants have applied for the job recently posted 136, interest in this sector is extremely healthy.

There is room for a number of players in this sector, we are so advanced in the Edge AI domain, that others will find the path to success
is going to be equally as hard, but there is no denying we are the company whose been facing ALL THE HEADWINDS out front for at
least 5 years, but I still remain very confident that we will succeed....roll on 2025.

Regards to all..........Tech.

 

[DJM263]

https://www.investors.com/news/technology/artificial-intelligence-stocks/

I like the part about:
"Many U.S. companies are pursuing custom AI software development projects, which will take longer to ramp up commercially"

 

[Space Cadet]

See if you can unwrap this. Latest presentation from Renesas in July. More confused than ever.

Dodgy, please help!

https://www.digitimes.com.tw/Semina...53AD07030/053AD07030_SGWXSG7FC1039CI771CB.pdf

SC

 

[IloveLamp]

Shelf Awareness: Tracking Inventory with AI

By using Edge Impulse, a highly accurate vision-based inventory tracking system can be produced very inexpensively, and without disrupting any existing business processes.

www.edgeimpulse.com

Keeping track of inventory can be daunting, but the latest advancements allows businesses to now leverage AI to streamline the process and boost efficiency. | Edge Impulse (a Qualcomm company)

Keeping track of inventory can be daunting, but the latest advancements allows businesses to now leverage AI to streamline the process and boost efficiency. Visit our new blog post to dive into how one sample edge AI-powered solution, using Raspberry Pi and BrainChip tools, can transform...

www.linkedin.com

 

[Space Cadet]

See if you can unwrap this. Latest presentation from Renesas in July. More confused than ever.

Dodgy, please help!

https://www.digitimes.com.tw/Semina...53AD07030/053AD07030_SGWXSG7FC1039CI771CB.pdf

SC

More from Renesas

https://www.renesas.com/us/en/about...ware-defined-vehicle-development-market-ready


SC

 

[supersonic001]

This is great Chippa !

May it be posted 1000 times over

Keeping track of inventory can be daunting, but the latest advancements allows businesses to now leverage AI to streamline the process and boost efficiency. | Edge Impulse (a Qualcomm company)

Keeping track of inventory can be daunting, but the latest advancements allows businesses to now leverage AI to streamline the process and boost efficiency. Visit our new blog post to dive into how one sample edge AI-powered solution, using Raspberry Pi and BrainChip tools, can transform...

www.linkedin.com

 

[AI_Inquirer]

I’m going to follow the example of a few others and leave the forum. Here’s what I would say if someone asked me what's the go BRN:

They are a small tech company aiming to introduce SNN’s to the market. Their technology is well timed and well positioned for the emerging edge market. The company operates on an IP based sales business model.

They launched the original version (1.0), which initially gained traction through license sales to Megachip and Renesas, as well as a joint development agreement (JDA) with Valeo. A batch of chips was produced, and the stock even rose above $2 when Mercedes announced its use in their new experimental prototype car. However, progress eventually stalled, and the company stated that the offering was too narrow and that good enough existing alternatives, made customers reluctant. Currently, neither Megachip nor Renesas have advanced 1.0 to productisation, so royalty revenue has not flowed. The CEO notes that there is still ongoing engagement for version 1.0.

The company continued to develop what they called the AKIDA1500. In August 2023, they received a shipment of AKD1500 chips in silicon from GlobalFoundries, utilising the latest SOI tech. These chips were distributed to customers as reference boards.

About a year later, they announced the early access availability of their second-generation AKIDA 2.0. They decided not to produce in silicon at this stage, although this could change in the future. Although there have been no commercial sales yet, the company continues state there is engagement actively and is still involved in evaluations with serious companies that have productisation deadlines.

Throughout the process, there have been numerous partnerships and the development of an ecosystem. Left over chips from the 1.0 chip run, have gone into edge boxes, to get SNN technology out there.

TENN is another technology the company plans to commercialise in the future. They are currently aggressively promoting it to build market awareness and funds raised from the latest CR will be directed toward its further development. As confirmed in our last price-sensitive announcement the SPP and the recent investor podcast, the CEO explained that they’re only now reaching the stage where they can offer the algorithm and are still building the pre-trained models, with ongoing development underway and future commercialisation to come.

BRN does not provide detailed updates or roadmaps on its commercialisation efforts, which requires investors to have a significant amount of faith. In the absence of company supplied information, investors have turned to two different forums for insights. On both forums, individuals who must spend over 20 hours a week posting have become so emotionally invested or financially trapped in the stock, so they aggressively counter any dissenting views, regardless of how absurd or misleading their claims may be. For instance, recent posts have suggested that TENN’s over-the-air software updates with Mercedes and Valeo are imminent, based on the use of the term "neural network" in LinkedIn posts and claimed that we have a product lineup with 14 offerings.

If you're interested in BRN, DYOR. Understand our current commercial offerings: AKIDA 1.0, AKIDA 1500, and AKIDA 2.0 IP, as well as our revenue model, including IP licensing and royalties. Examine the cash burn rate, cash runway, current revenue and the time likely needed to achieve profitability through IP sales and royalties. We are at a critical juncture.