BRN Discussion Ongoing

[manny100]

Manny through the looking glass ...

Better to see the blue sky than the grey clouds every day of the week.

 

[Humble Genius]

Could it be would it be can it be

View attachment 76867

Our forthcoming all-electric CLA has a remarkable potential range of more than 750 𝐤𝐦 in the WLTP. This is thanks to a huge number of painstaking efficiency measures.

One major contributor to the amazing consumption of just around 12 𝐤𝐖𝐡/100 𝐤𝐦 is the new Electric Drive Unit (EDU 2.0). We developed this in-house based on the groundbreaking technology of the VISION EQXX.

The main drive at the rear axle incorporates an innovative 2-𝐬𝐩𝐞𝐞𝐝 𝐠𝐞𝐚𝐫𝐛𝐨𝐱, which marks a 𝐟𝐢𝐫𝐬𝐭 𝐢𝐧 𝐭𝐡𝐞 𝐞𝐧𝐭𝐫𝐲-𝐥𝐞𝐯𝐞𝐥 𝐬𝐞𝐠𝐦𝐞𝐧𝐭. It combines dynamic response with excellent efficiency.

The first gear has a short 11:1 ratio for impressive off-the-line acceleration, high towing capacity and outstanding efficiency in urban traffic. The second gear, with its longer 5:1 ratio, is engineered for optimum power delivery at high speeds and for exceptional motorway efficiency.

The shift point between the two gears is dependent on the situation and the selected driving programme. It is continually optimised based on real-time parameters such as battery SoC, performance and driver requirements.

The driver, of course, senses none of this. Instead, they can simply enjoy seamless electric performance for miles on end.

"One major contributor to the amazing consumption of just around 12 𝐤𝐖𝐡/100 𝐤𝐦 is the new Electric Drive Unit (EDU 2.0). We developed this in-house based on the groundbreaking technology of the VISION EQXX."



OK, who's gonna have the honours of asking the question on Linkedin?

 

[Guzzi62]

Better to see the blue sky than the grey clouds every day of the week.

I think he meant to hint that you got it the wrong way around?

We do hardware and they do software, and they could like benefitting using our hardware.

 

[Pom down under]

 

[Pom down under]

I have never read so many positive comments on Tony Dawe and our CEO and board of management on hotcrapper as today,oh management are listening to shareholders, Yeah rightI like to donate $100.00 to the old bloke who abused them at the last AGM, Just so he goes to the next AGM, They will probably move it to the USA so as not to cop a earful. LISTEN THERE NOT LISTENING,Pigs in the trough that’s all they are. What the financials yep we are watchin Sean where the fk are they???? Are you happy clappers still clapping??
And the best part of the 4c was of course, the same old bullshit of a Defence contractor contract imminent to be signed just to string the happy clappers along some more. Shareholder ain’t listen to bullshit no more look forward to AGM to vote these furs out.

 

[Iseki]

I think he meant to hint that you got it the wrong way around?

We do hardware and they do software, and they could like benefitting using our hardware.

Nah, he doesn't know what he thinks. Hopefully we will see a new CEO, and he will like the new one just the same. So why not change? Not sure why relentlessly +ve posters never want change. They will be even more excruciating. The 4C is a debacle. It's the same as a re-rating of revenue. . We can't give akida away. If you can't see it, please stop posting how happy we should all be and maybe try a little less prozac.

 

[Galaxycar]

Pom with your pissy little investment I wouldn’t care either. What did you loose today $200.00

 

[Pom down under]

Pom with your pissy little investment I wouldn’t care either. What did you loose today $200.00

$15 to be precise

 

[manny100]

I think he meant to hint that you got it the wrong way around?

We do hardware and they do software, and they could like benefitting using our hardware.

Oops i knew that but somehow knackered the way it came out.
Rushing causes stuff ups.

 

[DK6161]

I have never read so many positive comments on Tony Dawe and our CEO and board of management on hotcrapper as today,oh management are listening to shareholders, Yeah rightI like to donate $100.00 to the old bloke who abused them at the last AGM, Just so he goes to the next AGM, They will probably move it to the USA so as not to cop a earful. LISTEN THERE NOT LISTENING,Pigs in the trough that’s all they are. What the financials yep we are watchin Sean where the fk are they???? Are you happy clappers still clapping??
And the best part of the 4c was of course, the same old bullshit of a Defence contractor contract imminent to be signed just to string the happy clappers along some more. Shareholder ain’t listen to bullshit no more look forward to AGM to vote these furs out.

Lol. Always funny when shareholders lost their sh!t

 

[Galaxycar]

Yep your all still clapping, bunch of flat earth believers.How does it look to have our management continuesly bagged out on one of the biggest forum im sites in Australia, the difference is until management and TD actually announce something that has revenue greater than a hotdog stand,it’s not hared to start to believe their incompetence

 

[DK6161]

Yep your all still clapping, bunch of flat earth believers.How does it look to have our management continuesly bagged out on one of the biggest forum im sites in Australia, the difference is until management and TD actually announce something that has revenue greater than a hotdog stand,it’s not hared to start to believe their incompetence

What are you talking about mate, TD and Pia are the 2 most hardworking employee and director that we have. Bonus shares are the least that we can give them. God bless them, and God bless brain chip Ikidyounot.

 

[Deadpool]

Yep your all still clapping, bunch of flat earth believers.How does it look to have our management continuesly bagged out on one of the biggest forum im sites in Australia, the difference is until management and TD actually announce something that has revenue greater than a hotdog stand,it’s not hared to start to believe their incompetence

Lets put this in a bit of perspective.
We all know it was a pretty ordinary report from BRN today, and unfortunately on the back of the Deepseek fever.
The collapse of Nvidia’s share price minus US$600 Billion and the “collateral damage” has been “enormous”.
Person wealth of
Jensen Huang minus 20.8 Billion as of today
Larry Ellison - 27.8 Billion today
Michael Dell -12.4 Billion today
Chengpeng zhao 12.1 Billion today
Even our own Greg Goodman losing $180 Million today.

Even if the 4c was great, the SP still would have tanked today.

 

[AusEire]

Pom with your pissy little investment I wouldn’t care either. What did you loose today $200.00

There's real small dick energy with this post

 

[Guzzi62]

Nah, he doesn't know what he thinks. Hopefully we will see a new CEO, and he will like the new one just the same. So why not change? Not sure why relentlessly +ve posters never want change. They will be even more excruciating. The 4C is a debacle. It's the same as a re-rating of revenue. . We can't give akida away. If you can't see it, please stop posting how happy we should all be and maybe try a little less prozac.

Nope, I post whatever I like, dude.

As long as I follow the general posting rules, nothing you can do about but put me on ignore.

 

[Diogenese]

I don't see DeepSeek as having a negative impact on BRN's business, although it may take a while for the ASX to twig.

It is hard to get your head around the tech generally, let alone the nuances which differentiate the various players.

However, there are largish differentiators between Akida and DeepSeek, edge/cloud, software/silicon, training v inference ... .

That's not to say we can rest on our laurels (do laurels have thorns?) - there is allways the chance that a competitor could spring from the bushes. However, our continuous stream of technological advances will help to ware off that contingency.

We must avoid becoming predicatble ... all that stuff about eternal vigilance:
View attachment 76865

Just on the training v inference dichotomy:

TRAINING:

It seems that DeepSeek compiles its models from other open source models.

Akida can also start with open source mdels and build on them. Akida does training really well with Edge Impulse.

INFERENCE:

DS uses software for inference.

Akida uses silicon for inference.

Now here's an analogy about use cases:

Think of the model as the library. In an initial phase, a lot of books are ordered and put on the shelves. This is equivalent to training.

As new books are added, this is equivalent to learning.

Then the people come to the library to obtain information by retrieving the relevant books and extracting the desired information. This can be thought of as inference.

This analogy is intended to show that the training phase of the model (stocking the shelves) is a large initial outlay. The adding of new books (learning new examples) can be an ongoing process, but the cost is small compare with training.

The main purpose of the library/model is to provide a resource for the users to carry out inference. Inference is the purpose of the model. Inferene is the purpose people use AI.

We know that Akida together with EI does training very well using N:M coding for optimal sparsity, as it does with learning (ML). Akida also does inference optimally using SNN, and N:M.

I don't know anything of how DS does inference other than it does it in software, and in the absence of any reference to "software secret sauce", I'll assune it's bog standard CNN, and we know that Akida eats software CNN.

So, in my view, the part of the SP fall today due to DS is misguided. If anything, this will boost Akida/TENNs because Akida/TENNs beats DS hands down at inference.

The caution on this is that, Akida does not do GenAI (analysis/reasoning), but for all the practical applications which do not require analysis/reasoning, Akida is superior.

That said, Akida can be used in conjunction with analysis/reasoning software and can relieve the software of much of the processing in cases like NLP, object tracking, etc., where long skip can play a part.

 

[AusEire]

Just gonna leave this here.

 

[Frangipani]

The points I raised suggest to me that Mercedes-Benz are still weighing their options. This doesn’t necessarily mean we are no longer engaged with them. Yet, it is an indisputable fact we are not the only ones in the neuromorphic space presently engaged with them.

What we can say for sure, though, is that not all of those engagements will ultimately translate into the signing of an IP license.
And so we wait…

View attachment 76816


The intelligent vehicle functionalities of the future call for pioneering new algorithms and hardware.

That’s why Mercedes-Benz is researching artificial neural networks that could radically increase the speed and energy efficiency of complex AI computations.

This could revolutionise future driving assistance and safety systems by overcoming the limits of today’s computing hardware.

For instance, conventional computing requires up to 3000 watts for advanced automated driving functions.

In future, neuromorphic computing could do this with just 300 watts, by mimicking the human brain with so-called “event-based processing“.

View attachment 76817 View attachment 76818 View attachment 76819 View attachment 76820

The following LinkedIn comment by Mercedes-Benz suggests to me once again that they continue to weigh their options with regards to various neuromorphic offerings - the good news is, we are apparently still in the running, but at the same time it is definitely not a confirmation of an upcoming deal either or an acknowledgment of Akida’s superiority over competitors’ technology (note they did not give Kimberly’s earlier post a ).

A very diplomatic reply, I’d say, focusing on the potential of BrainChip being considered in the future based on a positive past experience.
But no promise.

Mercedes-Benz AG on LinkedIn: Neuromorphic Computing

The intelligent vehicle functionalities of the future call for pioneering new algorithms and hardware. That’s why Mercedes-Benz is researching artificial…

www.linkedin.com

 

[miaeffect]

Just on the training v inference dichotomy:

TRAINING:

It seems that DeepSeek compiles its models from other open source models.

Akida can also start with open source mdels and build on them. Akida does training really well with Edge Impulse.

INFERENCE:

DS uses software for inference.

Akida uses silicon for inference.

Now here's an analogy about use cases:

Think of the model as the library. In an nitial phase, a lot of books are orederd ans put on the shelves. This is equivalent to training.

As new books are added, this is equivalent to learning.

Then the people come to the library to obtain information by retrieving the relevant books and extracting the desired information. This can be thought of as inference.

This analogy is intended to show that the training phase of the model (stocking the shelves) is a large initial outlay. The adding of new books (learning new examples) can be an ongoing process, but the cost is small compare with training.

The main purpose of the library/model is to provide a resource for the users to carry out inference. Inference is the purpose of the model. Inferene is the purpose people use AI.

We know that Akida together with EI does training very well using N:M coding for optimalparsity, as it does with learning (ML). Akida also does inference optimally using SNN, and N:M.

I don't know anything of how DS does inference other than it does it in software, and in the absence of any reference to "software secret sauce", I'll assune it's bog standard CNN, and we know that Akida eats software CNN.

So, in my view, the part of the SP fall today due to DS is misguided. If anything, this will boost Akida/TENNs because Akida/TENNs beats DS hands down at inference.

The caution on this is that, Akida does not do GenAI (analysis/reasoning), but for all the practical applications which do not require analysis/reasoning, Akida is superior.

That said, Akida can be used in conjunction with analysis/reasoning software and can relieve the software of much of the processing in cases like NLP, object tracking, etc., where long skip can play a part.

I wish your first name were Elon or Jenson

 

[DK6161]

View attachment 76875

Just gonna leave this here.

Ohhhh Mercedes, you are such a fking tease!