BRN Discussion Ongoing

[Guzzi62]

Eric Feuilleaubois (Ph.D) on LinkedIn: Comprehensive comparison between vision transformers and convolutional…

Comprehensive comparison between vision transformers and convolutional neural networks for face recognition tasks - Scientific Reports #DL #AI #ML…

www.linkedin.com

CNN VS ViTs
Looks like of late brainchip is comparing it's technology with existing technologies and shows the benefit.
Dyor

That paper isn't written by BrainChip, why do you think that?

ViT seems best for face recognition tasks, they write in conclusion.

It's written by:

Authors and Affiliations​

1. Grupo de Tratamiento de Imágenes (GTI), Information Processing and Telecommunications Center (IPTC), Universidad Politécnica de Madrid (UPM), Madrid, 28040, Spain
   Marcos Rodrigo, Carlos Cuevas & Narciso García

Comprehensive comparison between vision transformers and convolutional neural networks for face recognition tasks - Scientific Reports

This paper presents a comprehensive comparison between Vision Transformers and Convolutional Neural Networks for face recognition related tasks, including extensive experiments on the tasks of face identification and verification. Our study focuses on six state-of-the-art models: EfficientNet...

www.nature.com

 

[DJM263]

Have we seen anything from Spritle before?



Sorry if previously posted!

 

[Gemmax]

Data center emissions probably 662% higher than big tech claims. Can it keep up the ruse? — Guardian Australia

Emissions from in-house data centers of Google, Microsoft, Meta and Apple may be 7.62 times higher than official figures

apple.news

 

[Shadow59]

Data center emissions probably 662% higher than big tech claims. Can it keep up the ruse? — Guardian Australia

Emissions from in-house data centers of Google, Microsoft, Meta and Apple may be 7.62 times higher than official figures

apple.news

Here's an interesting idea. You know how we all hear these sci-fi stories about AI wiping out mankind because of the damage it's doing to the world. How about AI committing suicide after finding out what it's doing.

 

[Gazzafish]

Me:

what customers of megachips are using megachips ASIC solutions incorporating akida?

ChatGPT said:

MegaChips has incorporated BrainChip's Akida technology into its ASIC (Application-Specific Integrated Circuit) solutions. This partnership is enabling the development of next-generation edge AI technologies, especially in sectors like automotive, IoT, industrial robotics, and smart devices. MegaChips works with market leaders across various industries, but specific customer names have not been disclosed. The applications of these AI-powered ASIC solutions include home automation, security cameras, autonomous vehicles, medical devices, and gesture control, among others.

Me: I like it

 

[7für7]

Me:

what customers of megachips are using megachips ASIC solutions incorporating akida?

ChatGPT said:

MegaChips has incorporated BrainChip's Akida technology into its ASIC (Application-Specific Integrated Circuit) solutions. This partnership is enabling the development of next-generation edge AI technologies, especially in sectors like automotive, IoT, industrial robotics, and smart devices. MegaChips works with market leaders across various industries, but specific customer names have not been disclosed. The applications of these AI-powered ASIC solutions include home automation, security cameras, autonomous vehicles, medical devices, and gesture control, among others.

Me: I like it

as long as people take it as a fun, it’s all good. Let us not forget that ChatGPT take it’s informations from different sources and mix it up and create a good read… but if you ask “are you sure about that?” ChatGPT will respond “you are right thank you to pointing out my mistake… of course blablabla”

This is the response of my question if he would agree on this statement

“The statement seems generally reasonable and informative but does raise a few points worth considering:

1. Lack of Specific Customer Names: The mention that MegaChips works with market leaders across various industries without disclosing specific customer names could be seen as vague. It would be more credible if some examples were provided, though there could be legitimate reasons for confidentiality.
2. General Scope of Applications: While the sectors mentioned (automotive, IoT, etc.) and the applications (autonomous vehicles, gesture control, etc.) are plausible, the wide scope might appear a bit too broad unless more concrete details are provided. It might leave some readers questioning whether all of these applications are equally well-supported by the partnership.
3. Buzzwords: Terms like “next-generation edge AI” and “AI-powered ASIC solutions” are common in tech marketing. While not necessarily questionable, it would strengthen the statement if more specifics were provided about what sets these solutions apart from competitors’ technologies.

Overall, the statement is plausible but would benefit from more specifics to avoid skepticism.”

 

[Gazzafish]

as long as people take it as a fun, it’s all good. Let us not forget that ChatGPT take it’s informations from different sources and mix it up and create a good read… but if you ask “are you sure about that?” ChatGPT will respond “you are right thank you to pointing out my mistake… of course blablabla”

This is the response of my question if he would agree on this statement

“The statement seems generally reasonable and informative but does raise a few points worth considering:

1. Lack of Specific Customer Names: The mention that MegaChips works with market leaders across various industries without disclosing specific customer names could be seen as vague. It would be more credible if some examples were provided, though there could be legitimate reasons for confidentiality.
2. General Scope of Applications: While the sectors mentioned (automotive, IoT, etc.) and the applications (autonomous vehicles, gesture control, etc.) are plausible, the wide scope might appear a bit too broad unless more concrete details are provided. It might leave some readers questioning whether all of these applications are equally well-supported by the partnership.
3. Buzzwords: Terms like “next-generation edge AI” and “AI-powered ASIC solutions” are common in tech marketing. While not necessarily questionable, it would strengthen the statement if more specifics were provided about what sets these solutions apart from competitors’ technologies.

Overall, the statement is plausible but would benefit from more specifics to avoid skepticism.”

Very true. It was just some fun

 

[Diogenese]

Now that Sean has revealed that BRN has added a new product line in the form of algorithms (software), it would be nice to see the company make some noise about the introduction of the new product.

Software is a much quicker route to market, and it opens up an entirely new customer base. You would think that a product as snazzy as TeNNs software would substantially increase (double?) the company's market value. Of course, it's too soon to put an actual value on the product, so an ASX announcement is out. Does this mean we have to wait for an EAP to emerge from the NDA information vacuum?

Who will be first?

 

[IloveLamp]

Ramon.Space UK Awarded Program to Develop Advanced Onboard Communication Processors - RAMON.SPACE

Surrey, UK – 11th September 2024 – Ramon.Space, a global leader in space-resilient computing systems, proudly announces its selection for a groundbreaking program to develop next-generation digital onboard processors for large-scale communication constellations. This initiative is underpinned by...

ramon.space

Ramon.Space on LinkedIn: #spaceinnovation #ukspace #digitalpayloads #satellitetechnology | 13 comments

We’re proud to announce that Ramon.Space UK has been awarded a major program to develop next-gen onboard communication processors for large-scale communication… | 13 comments on LinkedIn

www.linkedin.com

 

[HopalongPetrovski]

Now that Sean has admitted that BRN has added a new product line in the form of algorithms (software), it would be nice to see the company make some noise about the introduction of the new product.

Software is a much quicker route to market, and it opens up an entirely new customer base. You would think that a product as snazzy as TeNNs software would substantially increase (double?) the company's market value. Of course, it's too soon to put an actual value on the product, so an ASX announcement is out. Does this mean we have to wait for an EAP to emerge from the NDA information vacuum?

Who will be first?

Hi Dio and other players.
What would you consider to be the likely/typical path to revenue for our snazzy TeNNs software?
i.e..... Do we licence its use for a particular product/family of products or get paid based upon the number of such products sold?

Beyond an influx of unaccounted for revenue are we likely to know that TeNNs has been used in anything commercial?
I'm assuming early adopters may want to keep this particular secret sauce as confidential as they can for all the same reasons that apply to silicon.

I recall you saying previously that you liked the idea of ongoing revenue from software updates which would occur much more frequently than upgrades to hardware.

Would just love some tangible progress that shifts the needle on our share price as would I'm sure most genuine holders. Hallelujah!!!
Come on BrainChip.
Coming up to my Ninth anniversary holding and Daddy wants a boat!

 

[Diogenese]

Hi Dio and other players.
What would you consider to be the likely/typical path to revenue for our snazzy TeNNs software?
i.e..... Do we licence its use for a particular product/family of products or get paid based upon the number of such products sold?

Beyond an influx of unaccounted for revenue are we likely to know that TeNNs has been used in anything commercial?
I'm assuming early adopters may want to keep this particular secret sauce as confidential as they can for all the same reasons that apply to silicon.

I recall you saying previously that you liked the idea of ongoing revenue from software updates which would occur much more frequently than upgrades to hardware.

Would just love some tangible progress that shifts the needle on our share price as would I'm sure most genuine holders. Hallelujah!!!
Come on BrainChip.
Coming up to my Ninth anniversary holding and Daddy wants a boat!

Hi Hoppy,

Good question. I hadn't really thought about the revenue model, and I don't know the basis on which production software is licensed, so I like your suggestion of a piece rate royalty.

It's not the same as MS Office or a manufacturing machine control software where the software is used to produce other documents or products.

When you think about it, companies like Valeo and Mercedes (if they are in fact using TeNNs/Akida software)would provide a copy of the software with their lidar/radar/E Class/DMS/NLP etc., so I suppose that a royalty will be paid for each copy of the software installed in a vehicle, and let's not forget the ongoing maintenance fees.

I don't think we will be doing mass market software in the near term, so there will be a substantial sign on licence fee.

 

[Slade]

Now that Sean has admitted that BRN has added a new product line in the form of algorithms (software), it would be nice to see the company make some noise about the introduction of the new product.

Software is a much quicker route to market, and it opens up an entirely new customer base. You would think that a product as snazzy as TeNNs software would substantially increase (double?) the company's market value. Of course, it's too soon to put an actual value on the product, so an ASX announcement is out. Does this mean we have to wait for an EAP to emerge from the NDA information vacuum?

Who will be first?

I’m going with Renesas as being the first to market with TeNNs.

 

[HopalongPetrovski]

Hi Hoppy,

Good question. I hadn't really thought about the revenue model, and I don't know the basis on which production software is licensed, so I like your suggestion of a piece rate royalty.

It's not the same as MS Office or a manufacturing machine control software where the software is used to prodce other documents or products.

When you think about it, companies like Valeo and Mercedes (if they are in fact using TeNNs/Akida software)would provide a copy of the software with their lidar/radar/E Class/DMS/NLP etc., so I suppose that a royalty will be paid for each copy of the software installed in a vehicle, and let's not forget the ongoing maintenance fees.

I don't think we will be doing mass market software in the near term, so there will be a substantial sign on licence fee.

Thanks for your thoughts.
Perhaps it's something like the right's associated with the listening to music.
The author/performer authorises an Apple or Spotify or other streaming service to make their song available and then receives a tiny amount each time the song is accessed by one of their customers.
This may involve extensive publicity commitments (trade show attendances) undertaken for a fee, at very little above cost, by the performer, in an attempt to expose our offering to a wide audience.
It used to be the music companies and more lately the entertainment conglomerates who produced the hard copies that were physically sold, and who then paid the artists, on a more or less, "piece rate".
The more popular the song, the more money produced and eventually distributed.
Now calculated by download quantities over a given period of time.

Additionally film/TV/game usage rights would be negotiated usually for a set fee upfront, based I guess, upon the balance of the need for usage by the film producer and the demand for the product experienced by the current corporate owner.
No matter how popular the film/TV show or game was, the provider of the component of the product (the song) just received that upfront set amount.
In some cases residual fees may be negotiated in the event of a long shelf life or some peculiarity which extended a particular songs longevity or periodic commemoration.

Eventually the copyright/patent period elapses or converts to a free use for everyone scenario and the pecuniary advantage lapses and is perhaps replaced with a mere acknowledgement to the originators contribution.

At the moment we are still that struggling artist looking for a Tata or similar to pick up, produce and promote the heck out of our song.

 

[Quiltman]

Drove to work with my son (25) this morning.
We are both BRN shareholders, his holding while small(ish), is significant to him.

What's going on with BrainChip he asked, is their tech any good ?

My answer, big picture view ... the conversation went something like this.

If you have truly disruptive, innovative tech ... who do you think would show interest, and in what order.

Pre anything - Evangelists, blue sky thinkers, and evangelist shareholders ( include myself in this group ).

First,
- Universities, research institutes ... a whole bunch of professors, PHD's and boffins running comparative studies, and publishing papers, demonstrating clear advantages of the tech.

Second,
- Something like space research and applications. Looking for highly disruptive technologies, high value, high intellect applying the tech - not consumer products ( far to risky )

Two big ticks there.

Next ... and Third,

- most likely a company/companies that have been working with the tech for several years, and have developed a high level of knowledge, trust and comfort. The tech is disruptive, but almost by definition also dangerous ( what happens if we get it wrong ! ). Start with one or two categories that this company has identified as large opportunities where the tech delivers a distinct advantage.

- for me , this sounds like TATA ( medical applications/ industrial ), and possibly Merc.

Fourth,

The company above ( TATA ? ) having successfully delivered proof of concept to mass market, encourages those companies that have been developing similar tech & applications with the confidence to follow.

Fifth,

The floodgates open.


....

So, I actually think we are on track.
The timelines have not met the expectations of many, but I believe the path to commercialisation is a predictable one - and we are on it.

Hopefully, before the end of the year, TATA has shown their hand and commitment.

So, I said to my son, tech is fine, let the process do it's thing.

 

[Diogenese]

Here's an interesting idea. You know how we all hear these sci-fi stories about AI wiping out mankind because of the damage it's doing to the world. How about AI committing suicide after finding out what it's doing.

To adapt the words of Sir Edward Coke in the case of Sutton's Hospital:

https://en.wikipedia.org/wiki/Case_of_Sutton's_Hospital

And it is great reason that an Artificial Intelligence in expectancy or intendment, or nomination, shall be sufficient to support the name of an Incorporation, when the Artificial Intelligence itself is onely in abstracto, and resteth onely in intendment and consideration of the model; for an Artificial Intelligence aggregate of many is invisible, immortal, & resteth only in intendment and consideration of the model; and therefore[1] cannot have predecessor nor successor.[2] They may not commit treason, nor be outlawed, nor excommunicate, for they have no souls, neither can they appear in person, but by programmer.[3] An Artificial Intelligence aggregate of many cannot do fealty, for an invisible body cannot be in person, nor can swear,[4] it is not subject to imbecilities*, or death of the natural body, and divers other cases.

* Really?

 

[Shadow59]

To adapt the words of Sir Edward Coke in the case of Sutton's Hospital:

https://en.wikipedia.org/wiki/Case_of_Sutton's_Hospital

And it is great reason that an Artificial Intelligence in expectancy or intendment, or nomination, shall be sufficient to support the name of an Incorporation, when the Artificial Intelligence itself is onely in abstracto, and resteth onely in intendment and consideration of the Law; for an Artificial Intelligence aggregate of many is invisible, immortal, & resteth only in intendment and consideration of the Law; and therefore[1] cannot have predecessor nor successor.[2] They may not commit treason, nor be outlawed, nor excommunicate, for they have no souls, neither can they appear in person, but by Attorney.[3] An Artificial Intelligence aggregate of many cannot do fealty, for an invisible body cannot be in person, nor can swear,[4] it is not subject to imbecilities*, or death of the natural body, and divers other cases.

* Really?

I can see you had a lot of fun with that!

 

[Smoothsailing]

What chance is BRN going into the Nintendo switch 2 ?.

 

[rgupta]

What chance is BRN going into the Nintendo switch 2 ?.

100-1

 

[Food4 thought]

100-1

Odd on favourite with those odds

 

[Space Cadet]

What chance is BRN going into the Nintendo switch 2 ?.

Even money. It either is or it isn't.

SC