BRN Discussion Ongoing

[Dhm]

Whilst we are considering our relationship with ARM it is good to remind ourselves of the massive revenue opportunities we have with them. This from ARM’s website about Brainchip and our best use cases…..

BrainChip

Ubiquitous, distributed AI.<br/>Close to the sensor.<br/>Inspired by the human brain. <br/>

www.arm.com

 

[Diogenese]

ISL Awarded a SBIR Phase III Contract Option | ISL - Technology for Real World Solutions

www.islinc.com

Just for interest.

"The contract continues the development of the B-1B Part-Task Trainer which started with radar scope interpretation and procedural training for the Offensive Systems Operator (OSO) and AFWERX “Accelerating Pilot to Combat-ready Aviators” task to demonstrate a F-15E application.

While Akida could do radar (oscillo)scope interpretation, I reckon Akida could interpret the radar signals before they are projected onto the screen.

 

[Diogenese]

That's a great article. Thanks for posting.

So as we are essentially an add on to an arm chip, I wonder what are we going to get per chip from arm? 3 cents? We are going to need tos sell a lot of chips!

I think Akida IP will be an optional extra at a price chosen by BrainChip.

 

[Build-it]

https://www.fierceelectronics.com/iot-wireless/arm-broadens-developer-access-neoverse-clouds

View attachment 21824

For a company that started 2022 by watching its pending acquisition by Nvidia fall apart, Arm has not spent any time looking back at what might have been. The firm has made numerous announcements in the last several months that appear aimed at proving its CEO’s pledge that the company has a “limitless future.”

The company also said its Scalable Open Architecture for Embedded Edge (SOAFEE) project to help create the software-defined automotive vehicle of the future has drawn 50 member companies since launching in September 2021. The project is due to offer up the second release of its reference implementation by the end of this year.

Mohamed Awad, SVP and GM, Infrastructure Line of Business.

He further explained that Works on Arm free developer access also is being offered through a partnership with miniNodes that supports hosted access to IoT and edge platforms, like Jetson Nano and Raspberry Pi, which support the development of edge native software and use cases.

Alot to like from this article for BRN shareholders and only supports how ARM needs Akida for its "limitless Future."

The move to an IP company is a master stroke imo and the royalties are the icing on the fiscal cake.

Edge Compute.

How's the numbers 15 mill !.

Arm’s focus will be primarily on hyperscaler, cloud, IoT, automotive and the Metaverse, he said, but added, “we have other things up our sleeves” to be revealed later. “That’s one of the fun things about Arm: there’s not a technology space that we cannot participate in.”

Arm’s earlier contributions to power-efficient chip designs for smartphones will have bearing on upcoming technologies, everything from data center to AR and VR glasses. “We are in the early days of AR and VR glasses which are compute intensive and need great battery life and a broad suite of apps, and that’s about a broad developer community,” Haas said.

Speaking of which, Arm counts 15 million developers in its software development ecosystem with 10 million Arms-based apps that run on every major OS.

Segars appears to have left Arm on good terms, saying in a statement, “I’m very bullish on Arm’s future success under Rene’s leadership and can’t think of anyone better to lead the company through its next chapter.”

While Segars spent 30 years at Arm, moving from graduate engineer to CEO, Haas has been president of the Arm IP Products Group since 2017 after first joining Arm in 2013. Haas said he plans to continue to reside in San Jose, California, with plenty of travel to Cambridge, England, where Arm is based.

And one for Dio...

Arm currently has 6,400 employees, including about 5,200 engineers. Asked if the headcount of engineers would change with the Arm’s focus on an IPO, Haas said, “you can never have enough engineers.”

Edge Compute.

 

[Iseki]

Lies…and bloody lies !!

At school I was proficient, and not a member of the A Forms. The only prize I ever collected on the numerous Speech Day‘s I attended was the Parents and Friends Essay prize. This encouraged me to write, and upon leaving school I was fortunate enough to win one of the few cadetships on The Advertiser…South Australia’s morning broadsheet.

In the late 60’s, budding journos didn’t go to uni to get a degree in communications or the like, they had to win a Cadetship from a media organisation. As with most apprenticeships, this entitled them to enjoy all the most rudimentary jobs. However, each cadet was assigned a mentor, and I was lucky enough to get Des, one of the ‘Tisers most highly regarded Features writers.

Des was a good mate of Max, a Contributing writer for the paper. Each Thursday evening they would get together at the Criterion, the paper’s watering hole. Occasionally they would be joined by a colleague from The News, Adelaide’s evening tabloid. I was invited to attend for the sole reason of fetching the next round, but it afforded me the opportunity to listen in on the lively conversation. Des and Max we’re old school journos, who did everything by the book…to only commit to paper something that was properly researched and accurate. Their colleague used to mock them for this, saying whatever it took to sell newspapers…his name was Rupert.

Like the SMH, the Age and other state morning broadsheets back then, The Advertiser was a highly successful family owned newspaper that dined out on classified advertising known as ‘the rivers of gold’. They were able to afford the best journos and writers, and allow them the time to produce well researched stories.

That era has passed, and as we’re all aware newspapers, in fact all traditional mainstream media are struggling to stay afloat. The fallout of this is poorly trained journos with little time for research, being pressured to present controversial stories…the more sensational the better to increase ratings or circulation in order to survive.

Added to this, they now have to compete with the totally unregulated, irresponsible, often defamatory world that is social media.

As Rupert would say to Des and Max…whatever it takes to sell newspapers.

If you can accept the above, it will help you understand why we get what we do from such esteemed financial publications as MF, the AFR et al. Of course there is no excuse for this, but don’t get your knicker’s in a twist… it’s just a sign of the times.

Btw…scalability has been raised in recent posts. You may find this hard to believe after I berated Lou and his fellow board members who attended the 2018 AGM, but we actually struck up a friendship of sorts. At a lunch we had, I recall him saying that the lack of scalability was a major reason why Studio was shelved.

What are you talking about? The 'Tiser was owned by Rupert's dad. You were given your start by the Murdochs.

 

[Fact Finder]

That's a great article. Thanks for posting.

So as we are essentially an add on to an arm chip, I wonder what are we going to get per chip from arm? 3 cents? We are going to need tos sell a lot of chips!

Well selling a lot of chips is the plan all tied up in past statements about saturating the market and become ubiquitous and essential but in all the numbers that have been played with we do not have what ARM is paying out to the 1,000 partners who are part of the ARM ecosystem.

All we have is ARMs profit from its operations. Profits do not include what the partners charge for use of their IP and software by ARM.

So it is impossible to extrapolate from ARMs profit what Brainchip will be charging for its IP when included in a package sold to an ARM customer.

If we go back to the first Pitt Street coverage of Brainchip they gave numbers and how AKIDA IP might end up being charged out at in the following extract:

“IP licencing fees: Once the NRE work is complete and customers want to move into commercial production, they typically pay a one-off licence fee for the use of the technology. License fees can vary broadly and will be different for each customer, depending on the intended application areas, expected production volumes, the amount of IP to be used, etc. License fees typically vary from $1m for single-product use to $3m for multi-product use. Non- exclusive licencing will help the company sign multiple licensees to expand its revenue-generating potential.
Royalties: We believe the most lucrative future revenue stream for the company will be royalties paid by customers for each product they sell that includes Akida IP. These royalties are usually a percentage of the customer’s revenue from sales and typically range from 2% to 15%, again depending on the intended application areas, the amount of IP used and expected production volumes. Notably, royalty percentages also depend on the uniqueness of the IP that is being licensed. As the specifications and features of Akida are quite unique vs. other technologies, including Intel’s Loihi and IBM’s TrueNorth, this may help the company charge higher-than-average royalty percentages for Akida.
Other royalty revenue models simply use a fixed dollar amount per chip sold. This is a preferred model for many high-volume production companies, including cell phone manufacturers”

https://static1.squarespace.com/static/5af533a312b13fb602fe7d7b/t/612313b3aae13c19f15e79bb/1629688762738/BrainChip+Holdings+research+initiation+report+-+20+08+2021.pdf

Remember the beauty of the Brainchip neural fabric is that it is made up of intelligent nodes. A customer can licence one node or as many as they need.

AKIDA 1.0 uses 80 nodes so in just this one chip there are effectively 80 different sales points or product offerings when selling the IP.

When AKIDA 1.0 was initially being discussed Peter van der Made and Co stated they could produce it to allow for 1,024 chips to be ganged together which is 80 x 1,024 = 81,920 nodes.

In the end they set it up to allow 64 which is 80 x 64 = 5,120 nodes.

In IP there are no limits on how big it can scale so the sales options are limitless.

From the far, far Edge to the data centre and beyond through the cloud into space AKIDA technology aims to become ESSENTIAL Ai.

My opinion only DYOR
FF

AKIDA BALLISTA

PS: As Peter van der Made and Anil Mankar and the band of brilliant engineers at the North Star in Perth bring AKIDA closer to AGi it becomes more and more unique and so the value of what can be charged for each new iteration of the IP increases. AKIDA 2.0 will close the gap to 15% left open by AKIDA 1.0.

 

[HopalongPetrovski]

Hiya HP,
Moved over here a long time ago. Observe and listen. HC is a waste of space.
Be well

Thanks Adam.
Only just became aware of your new moniker.
Please keep speaking your truth, yours is a voice of experience needed here.

 

[Fact Finder]

How's the numbers 15 mill !.

Arm’s focus will be primarily on hyperscaler, cloud, IoT, automotive and the Metaverse, he said, but added, “we have other things up our sleeves” to be revealed later. “That’s one of the fun things about Arm: there’s not a technology space that we cannot participate in.”

Arm’s earlier contributions to power-efficient chip designs for smartphones will have bearing on upcoming technologies, everything from data center to AR and VR glasses. “We are in the early days of AR and VR glasses which are compute intensive and need great battery life and a broad suite of apps, and that’s about a broad developer community,” Haas said.

Speaking of which, Arm counts 15 million developers in its software development ecosystem with 10 million Arms-based apps that run on every major OS.

Segars appears to have left Arm on good terms, saying in a statement, “I’m very bullish on Arm’s future success under Rene’s leadership and can’t think of anyone better to lead the company through its next chapter.”

While Segars spent 30 years at Arm, moving from graduate engineer to CEO, Haas has been president of the Arm IP Products Group since 2017 after first joining Arm in 2013. Haas said he plans to continue to reside in San Jose, California, with plenty of travel to Cambridge, England, where Arm is based.

And one for Dio...

Arm currently has 6,400 employees, including about 5,200 engineers. Asked if the headcount of engineers would change with the Arm’s focus on an IPO, Haas said, “you can never have enough engineers.”

Edge Compute.

Didn’t the nice Mr. Segars take a seat on the Board of Edge Impulse?

Just checked and yes he did:

Big News! Edge Impulse Announces Simon Segars Joining the Board of Directors

Simon Segars, the distinguished technology leader and former CEO of Arm, brings deep industry experience and leadership as he joins our board.

www.edgeimpulse.com

My (not so) vague memory only so DYOR
FF

AKIDA BALLISTA

 

[Fact Finder]

How's the numbers 15 mill !.

Arm’s focus will be primarily on hyperscaler, cloud, IoT, automotive and the Metaverse, he said, but added, “we have other things up our sleeves” to be revealed later. “That’s one of the fun things about Arm: there’s not a technology space that we cannot participate in.”

Arm’s earlier contributions to power-efficient chip designs for smartphones will have bearing on upcoming technologies, everything from data center to AR and VR glasses. “We are in the early days of AR and VR glasses which are compute intensive and need great battery life and a broad suite of apps, and that’s about a broad developer community,” Haas said.

Speaking of which, Arm counts 15 million developers in its software development ecosystem with 10 million Arms-based apps that run on every major OS.

Segars appears to have left Arm on good terms, saying in a statement, “I’m very bullish on Arm’s future success under Rene’s leadership and can’t think of anyone better to lead the company through its next chapter.”

While Segars spent 30 years at Arm, moving from graduate engineer to CEO, Haas has been president of the Arm IP Products Group since 2017 after first joining Arm in 2013. Haas said he plans to continue to reside in San Jose, California, with plenty of travel to Cambridge, England, where Arm is based.

And one for Dio...

Arm currently has 6,400 employees, including about 5,200 engineers. Asked if the headcount of engineers would change with the Arm’s focus on an IPO, Haas said, “you can never have enough engineers.”

Edge Compute.

So the Train Drivers Union was right. And I thought they were just protecting engineers jobs from the introduction of self driving trains.

 

[Sirod69]

Sorry I will listen tomorrow!

Here’s something for our German-speaking followers. It was my pleasure recently to sit down for a chat with Luca Leicht and Gerd Stegmaier for the MO/OVE podcast from auto motor und sport.

Future EV innovations, digital technology, software, our Mercedes-Benz strategy, the importance of collaboration and, of course, plenty of car talk across the performance spectrum. There’s a LOT going on at Mercedes-Benz AG right now, so we covered a LOT of ground.
Have a listen!

https://lnkd.in/eFD86rpU

Moove (94) – Markus Schäfer, CTO Mercedes Benz: Werden elektrische Mercedes teurer Luxus?

Hat die A-Klasse noch Platz in der Luxusstrategie? Wie kommt die Emotion in einen elektrischen AMG und wieso entwickelt Mercedes noch immer neue Plug-In-Hybride? Entwicklungschef Markus Schäfer erklärt im Podcast, wie sich der Autobauer für die Zukunft rüstet.

www.auto-motor-und-sport.de

 

[Sirod69]

I like it!


Edge Impulse

Researchers are leveraging AI technology to passively track whale sharks and to prevent unintentional bycatch of endangered, threatened, and protected species within the Galápagos Marine Reserve.

How A.I. is helping to protect the endangered whale sharks of the Galapagos

This technology can classify each whale shark based on the spot patterns on their bodies, which are unique like a fingerprint.

fortune.com

 

[Fact Finder]

Didn’t the nice Mr. Segars take a seat on the Board of Edge Impulse?

Just checked and yes he did:

Big News! Edge Impulse Announces Simon Segars Joining the Board of Directors

Simon Segars, the distinguished technology leader and former CEO of Arm, brings deep industry experience and leadership as he joins our board.

www.edgeimpulse.com

My (not so) vague memory only so DYOR
FF

AKIDA BALLISTA

I think it is worthwhile to have regard to what Edge Impulse and Simon Segars sees as its corporate strategy/mission:

“Simon shares our vision of enabling millions of developers to engineer with ML across the whole silicon industry, much like Arm did to enable compute. It’s a big honor to have him choose us for one of just a few board positions he has accepted, and is an amazing testament to what we’ve achieved with Edge Impulse.

“Unlocking our world's potential through smarter, more efficient, and accessible computing has been a consistent theme in my professional journey,” Simon says. “Edge Impulse shares this vision and is at the center of the next frontier of edge computing, by empowering developers to harness machine learning across all classes of edge devices.”

When you read about the past history of the founders of Edge Impulse with Simon Segars it is clear at least to me that they will be very open to listening to what he has to say and it is in this context that since his appointment in June, 2022 the adoption of and promotion by Edge Impulse of the Brainchip AKIDA Edge technology solution has been gaining pace.

In other words I think Simon Segars is onboard.

I am a great fan of ancient history and how cultures in for example ancient Egypt would pursue projects which to complete stretched over generations and hundreds of years.

Can you imagine today in so called modern Australia hearing a leader regardless of politics proposing an infrastructure project with a time line for completion of 300 years.

Of course you can’t but what is even more unimaginable is an electorate coming behind this leader and supporting their vision.

We marvel at things bequeathed to us like the pyramids in Egypt but today absolutely nothing is built to survive for eternity.

In three, four or five thousand years what will the legacy of our generation be that future generations will look back at in wonderment.

We are the instant gratification generations and in another 1,000 years when the Sydney Opera House is dust the Egyptian pyramids will still be here.

So what is the point I am making. It is pretty simple. The current economic and political situation will pass it is not even a speck on the lenses of time and the amazing ecosystem opportunities and partnerships discussed over the last 24 hours will make Brainchip into the successful corporation we desire.

Weighed against the patience and commitment displayed by our primitive ancient ancestors which extended over hundreds of years waiting a couple of years for Brainchip to emerge seems like absolutely nothing at all.

Waiting until the next AGM in May, 2023 to measure the CEO Sean Hehir’s achievements seems in this context something silly to be even discussing.

An ancient Egyptian would not think Sean Hehir much of a leader if his vision was to be measured in 12 monthly increments.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Baneino]

Maybe just a little thing and maybe already known. I noticed that we appear on the page of arm at partners under relevance at the top.
How is this to be evaluated ? Where I live, that means a lot!

 

[Fact Finder]

Maybe just a little thing and maybe already known. I noticed that we appear on the page of arm at partners under relevance at the top.
How is this to be evaluated ? Where I live, that means a lot! View attachment 21831

Hi @Baneino

Thank you for posting it is great to have your input.

I am not sure I understand your question could you expand on it please?

Regards
FF

AKIDA BALLISTA

 

[Falling Knife]

Hi @Baneino

Thank you for posting it is great to have your input.

I am not sure I understand your question could you expand on it please?

Regards
FF

AKIDA BALLISTA

FF May I translate for you
When you search for an ecosystem partner on the ARM website using the search "by relevance"
https://www.arm.com/partners/catalog/results#sort=relevancy

Brainchip is offered by Arm (Therefore potentially considered by ARM) as the first preference for that chosen solution.
Its not alpah order so the decision to have a particular company first in the offereings seems based on another reason.

In Germany being FIRST is seen to be an important statement
Mercedes knows this

 

[Fact Finder]

FF May I translate for you
When you search for an ecosystem partner on the ARM website using the search "by relevance"
https://www.arm.com/partners/catalog/results#sort=relevancy

Brainchip is offered by Arm (Therefore potentially considered by ARM) as the first preference for that chosen solution.
Its not alpah order so the decision to have a particular company first in the offereings seems based on another reason.

In Germany being FIRST is seen to be an important statement
Mercedes knows this

Thank you @Falling Knife it all becomes clear. Sorry for being obtuse I can only blame my adopting the modern thinking that even coming last deserves a prize.

As a reward I will share this which I just found yet another use for AKIDA technology. Hot of the academic press of 6.11.22:


“V. CONCLUSION
In this work we introduced an equalizer based on an SNN and inspired by the structure of the DFE. Furthermore, we pro- posed ternary encoding, which is a bipolar encoding based on a quantizer. We compared our proposed approach against linear equalizers, the DFE and ANN-based equalizers. Our proposed approach is able to execute equalization for various frequency- selective channels, clearly outperforming linear equalizers and with similar performance than the DFE and ANN-based methods. We furthermore showed that ternary encoding is a robust and fast encoding technique, that can outperform log- scale encoding. This work lays the fundamentals for future energy efficient communication receivers that use neuromor- phic hardware based on spikes that promise significantly lower energy consumption than traditional receiver circuits”

https://arxiv.org/pdf/2211.04756

As if Brainchip needed any other industry to dominate.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Diogenese]

Hi @Baneino

Thank you for posting it is great to have your input.

I am not sure I understand your question could you expand on it please?

Regards
FF

AKIDA BALLISTA

The options on the Sortieren nacht/Sort by drop-down box have BrainChip at the top of the "Relevance" list.

One explanation is that BrainChip is relevant to more of the categories ARM has allocated to the Partners, eg :

https://www.arm.com/partners/catalog/results#sort=relevancy

In other words, Akida can be implemented in more of these use cases than any of the other ARM partners, which has interesting implications for our discussion about the % of ARM products with which Akida IP will be used.

 

[Fact Finder]

The options on the Sortieren nacht/Sort by drop-down box have BrainChip at the top of the "Relevance" list.

One explanation is that BrainChip is relevant to more of the categories ARM has allocated to the Partners, eg :

View attachment 21848

https://www.arm.com/partners/catalog/results#sort=relevancy


View attachment 21846


In other words, Akida can be implemented in more of these use cases than any of the other ARM partners.

Your making out AKIDA to be like the try hard kid in my maths class who always had his hand up to answer questions and everyone hated.

 

[Diogenese]

The options on the Sortieren nacht/Sort by drop-down box have BrainChip at the top of the "Relevance" list.

One explanation is that BrainChip is relevant to more of the categories ARM has allocated to the Partners, eg :

View attachment 21848

https://www.arm.com/partners/catalog/results#sort=relevancy


View attachment 21846


In other words, Akida can be implemented in more of these use cases than any of the other ARM partners.

NXP are the next most versatile, but they are no competition - they would not know a neuromorph if they fell over it:

https://www.digikey.com.au/en/product-highlight/n/nxp-semi/i-mx-8m-nano-ultralite

 

[Fact Finder]

If you have a few spare minutes this little article is very illuminating when you know as we do that Brainchip has the first commercial neuromorphic IP and reference chip capable of:

1. On chip one shot, few shot and incremental learning;

2. On chip multi sensor processing - evidence Nviso working towards having AKIDA 1.0 process all 20 odd Apps on chip, and

3. On chip multi sensor fusion to create actionable instructions.

AI Uses Artificial Sleep To Learn New Task Without Forgetting The Last - Public News Time

Many AIs can only become good at one task, forgetting everything they know if they learn another. A form of artificial sleep could help stop this from

publicnewstime.com

As the concluding paragraph states it is great to think ahead as Peter van der Made and Anil Mankar have by at least three years over everyone else in the World protected by patents.

My opinion only DYOR
FF

AKIDA BALLISTA