BRN Discussion Ongoing

[Diogenese]

Applications that can be greatly accelerated by Helium are Fast Fourier Transform (FFT).

Now where did we hear the term Fast Fourier Transform before? Oh, that's right, the Ninteno patent.



View attachment 38531


View attachment 38530

Here's a little song I wrote.
I'm gonna sing it to you note by note:

A microphone has a membrane which reverberates in response to the air pressure variations caused by sound. Sound can be broken down to a whole series of sine waves (Fourier Transform). An ADC samples the instantaneous amplitude combined signal at a bit over twice the rated frequency. Higher frequencies are blocked by a low pass filter.

The speed of travel of sound is independent of frequency - a 1 kHz wave travels at the same speed as a 4 kHz wave (750 mph) you just get more waves in the same distance. This also explains the doppler effect. The wavelength is compressed by an approaching sound source, and stretched out by a receding source.

Similarly in a wireless phone, the speed of propagation is constant and independent of the frequency of the carrier signal.
Modulation in digital wireless is a bit more complex.

Assume we use 8-bit modulation to represent the instantaneous amplitude od the samples analog signals. We have a stream of 8-bit bytes made up of ones and zeros, but we can't send these bytes as is over RF. so we convert them to RF pulses - a few cycles of RF carrier for 1s and none for 0s.

The carrier frequency can be Mega Hz, the higher the frequency, the closer the bits can be packed. This allows for time compression of speech in packets of bytes at the transmitter. This means that the wireless speech packet only uses a small part of the time available on the carrier wavelength, so other message packets can be fitted into the same carrier. This is Time Division Multiplexing. Each packet has a label so it can be reassembled at the receiver with other packets from the same message.

Again, the speed of propagation of radio waves (186,000 mph) is independent of the frequency of the carrier wave.

At the receiver, the packets of the message are decompressed and reassembled into the digital bytes.

The digital bytes can then be converted to audio via a DAC to be fed to a speaker.

Don't worry,
Be happy!

 

[manny100]

Our Patent expansion worldwide with many, many Patents currently awaiting examination, if granted will narrow the channel even
further for current players and emerging players, the question is, how many ways are there to achieve what Peter, Anil and teams have
created to this point, if it was as easy as that, many companies would already be boasting that they have achieved Edge AI on a grand
scale, but all I hear is talk, we may be in the same category at this current point in time, but the IBM's, the Qualcomm's, the Intel's etc
are all in the same boat, who has the code to crack open the future, the race is on, are we still in front, who really knows.

This future pie will be shared between many, as in my opinion, many are already sleeping with the enemy, so to speak, as we have all
learnt, many companies have something to bring to the table, so letting go of egos, or power plays will smooth the runway to success
for all.

Brainchip has "proven" technology that the others don't, we can both benefit through mutual co-operation, keeping the doors open
will ultimately benefit us all.

That's why our growing partnership list is so important, it's the clients of clients of clients that ultimately builds a huge network of
solid trust between all parties concerned.

Thanks to all the dedicated researchers still contributing to this forum, I and many others appreciate all your efforts, cheers.

Tech

Agree, we certainly have a 2 plus year lead over other potential entrants who are still in research.
Once they think they have it right as Sean said recently its a long journey to then get it to market. So our lead could easily be 3 years.
That is our major risk.
We need to be in in commercial use and growing rapidly before we see genuine competition appear.
As Sean said once we are in a product we will be in for many future generations of that product.
I am confident that AKIDA will be in products being commercially available by year end or shortly thereafter with revenue to follow.
At the moment the market is reflecting the above risk in our current SP.
Markets are fueled by expectation and as we get closer to year end these expectations will be reflected in the SP.

 

[jtardif999]

I was watching last NVIDIA prestations on YouTube and Jensen was quite vocal about big processors and how green is new NVIDIA GPU is which consumes only 580 watts while the other consume 1 kWh.
NVIDIA is selling their business as data centres are new computers.
We are selling our business with micro to milliwatts of energy consumption but NVIDIA is still green coz they can keep all the raw data for you even if you need that or not??
To me greed is at its peak at the stake of environment.
Dyor

Patience will win out though, Nvidia can plug the crap out of their GPUs but ultimately GPUs are old tech and have a limited capacity in the ensuing AI future - moving away from large compute to doing most of it outside of the data centre. There’s so much hype around LLMs and Nvidia are cashing in on the compute required, but it is not a sustainable model and will need a revamp. I think this will kill Von Neumann quicker, actually as it will be excessive to maintain in its current form. On the other hand we are hiding at the edge supporting the infrastructure required to compute there and ultimately this is where the majority of compute (for LLMs and everything else) will go. AIMO.

 

[Tothemoon24]

You little beauty Dio

 

[Diogenese]

You little beauty Dio




View attachment 38532

Missed it by that much!

 

[manny100]

Interesting Mega Chips (MC) integrating AKIDA.
A customer of MC is Qualcomm who evidently supply to Prophesse.
At around the 11 min mark of the Understanding AI video interview (day after AGM) when talking about Tech partnerships Sean said
"if you want to work with a Prophesse camera you want to know it works well with Brainchip."
Its all slowly falling into place but Sean and his team cannot say much because of the NDAs.

 

[equanimous]

I was watching last NVIDIA prestations on YouTube and Jensen was quite vocal about big processors and how green is new NVIDIA GPU is which consumes only 580 watts while the other consume 1 kWh.
NVIDIA is selling their business as data centres are new computers.
We are selling our business with micro to milliwatts of energy consumption but NVIDIA is still green coz they can keep all the raw data for you even if you need that or not??
To me greed is at its peak at the stake of environment.
Dyor

NVIDIA GPUs are definitely not green because the nvidia gpu I use (350 watts) is enough to substitute not having a heater in my room on winter nights.

 

[Deleted member 118]

Here's a little song I wrote.
I'm gonna sing it to you note by note:

A microphone has a membrane which reverberates in response to the air pressure variations caused by sound. Sound can be broken down to a whole series of sine waves (Fourier Transform). An ADC samples the instantaneous amplitude combined signal at a bit over twice the rated frequency. Higher frequencies are blocked by a low pass filter.

The speed of travel of sound is independent of frequency - a 1 kHz wave travels at the same speed as a 4 kHz wave (750 mph) you just get more waves in the same distance. This also explains the doppler effect. The wavelength is compressed by an approaching sound source, and stretched out by a receding source.

Similarly in a wireless phone, the speed of propagation is constant and independent of the frequency of the carrier signal.
Modulation in digital wireless is a bit more complex.

Assume we use 8-bit modulation to represent the instantaneous amplitude od the samples analog signals. We have a stream of 8-bit bytes made up of ones and zeros, but we can't send these bytes as is over RF. so we convert them to RF pulses - a few cycles of RF carrier for 1s and none for 0s.

The carrier frequency can be Mega Hz, the higher the frequency, the closer the bits can be packed. This allows for time compression of speech in packets of bytes at the transmitter. This means that the wireless speech packet only uses a small part of the time available on the carrier wavelength, so other message packets can be fitted into the same carrier. This is Time Division Multiplexing. Each packet has a label so it can be reassembled at the receiver with other packets from the same message.

Again, the speed of propagation of radio waves (186,000 mph) is independent of the frequency of the carrier wave.

At the receiver, the packets of the message are decompressed and reassembled into the digital bytes.

The digital bytes can then be converted to audio via a DAC to be fed to a speaker.

Don't worry,
Be happy!

Nice poem FF

 

[Bravo]

Here's a little song I wrote.
I'm gonna sing it to you note by note:

A microphone has a membrane which reverberates in response to the air pressure variations caused by sound. Sound can be broken down to a whole series of sine waves (Fourier Transform). An ADC samples the instantaneous amplitude combined signal at a bit over twice the rated frequency. Higher frequencies are blocked by a low pass filter.

The speed of travel of sound is independent of frequency - a 1 kHz wave travels at the same speed as a 4 kHz wave (750 mph) you just get more waves in the same distance. This also explains the doppler effect. The wavelength is compressed by an approaching sound source, and stretched out by a receding source.

Similarly in a wireless phone, the speed of propagation is constant and independent of the frequency of the carrier signal.
Modulation in digital wireless is a bit more complex.

Assume we use 8-bit modulation to represent the instantaneous amplitude od the samples analog signals. We have a stream of 8-bit bytes made up of ones and zeros, but we can't send these bytes as is over RF. so we convert them to RF pulses - a few cycles of RF carrier for 1s and none for 0s.

The carrier frequency can be Mega Hz, the higher the frequency, the closer the bits can be packed. This allows for time compression of speech in packets of bytes at the transmitter. This means that the wireless speech packet only uses a small part of the time available on the carrier wavelength, so other message packets can be fitted into the same carrier. This is Time Division Multiplexing. Each packet has a label so it can be reassembled at the receiver with other packets from the same message.

Again, the speed of propagation of radio waves (186,000 mph) is independent of the frequency of the carrier wave.

At the receiver, the packets of the message are decompressed and reassembled into the digital bytes.

The digital bytes can then be converted to audio via a DAC to be fed to a speaker.

Don't worry,
Be happy!

That's a lovely little ditty Dodgy-Knees! Imagine how cool it would be if you could get Kylie Minogue to sing it. I'm pretty sure it would go straight to number 1 on the pop charts.

I just wish I knew what the lyrics actually meant.

 

[equanimous]

That's a lovely little ditty Dodgy-Knees! Imagine how cool it would be if you could get Kylie Minogue to sing it. I'm pretty sure it would go straight to number 1 on the pop charts.

I just wish I knew what the lyrics actually meant.

Kylie-Meknees

 

[rgupta]

Nice poem FF

Any reason for FF

 

[yogi]

This explains a lot to those who are getting anxious/upset and lack of revenue

Dirk Jan Swagerman on LinkedIn: #softwareengineering #systemsengineering #systemsthinking #software… | 13 comments

Well spend 120 seconds of your time. If you treat software engineering as a commodity that is best outsourced to suppliers (based on price), learn from the… | 13 comments on LinkedIn

www.linkedin.com

 

[Deleted member 118]

Any reason for FF

 

[TECH]

It definitely wasn't Fact Finders poem (Fact)

But it did sound like the type of response Peter would have written if asked a specific technical question from an engineer.

In my opinion of course....Tech

 

[Xray1]

Yes, but why is that and what will it take to get mainstream enthusiasm for BRN. Our team cleverly started positioning in the AI space well before the current buzz around AI yet we constantly appear to be overlooked. I hope this type of commentary does not reflect what our potential customers and industry participants think of AKIDA. Uptake of GEN 2 needs to be significant and rapid once it's fully released in the next couple of months. Looking forward to the next GEN 2 announcement.

I agree with your post .............. But imo, I suppose that all investors / shareholders and potential customers are essentially now waiting for the Akida 1500 and 2000 roll outs to be fully bench tested to make sure that they are fully functional and compliant to Co and end user specifications ... let's hope, that there a no hiccups along the way that Anil can't control.

 

[Yoda]

View attachment 38494

Does Renesas have any other AI partners or are we the only one? Putting that question another way, can we definitely conclude that this is us in their products? I'm certainly hoping so . . .

 

[Shadow59]

Just for a laugh, just now logged into Commsec and BRN

screen

 

[IloveLamp]

Mercedes-Benz Media

Exklusive Einblicke und individuelle Angebote: Erleben Sie mit Mercedes-Benz das Maximum aus digitaler Live-PR. Exclusive insights and individual offers: Experience the maximum of digital live PR with Mercedes-Benz.

media.mercedes-benz.com

 

[IloveLamp]

Does Renesas have any other AI partners or are we the only one? Putting that question another way, can we definitely conclude that this is us in their products? I'm certainly hoping so . . .

They have lots of ai partners...........or at least they did...

Why would they choose them over our ip though?

That is the genius in making this technology "cheap". We know we have the goods, cost of buying is not an objection now because we are competitively priced.



Developing the technology to integrate into future and existing products could be a road block, but clearly the technology is good enough and proven enough for certain forward thinking companies, and the rest will have no choice but to eventually follow suit.

Either they get left behind, or come up with something equally impressive or better.........

Good luck to whoever is in charge of that lol!

 

[Iseki]

Helium isn't hot air, but it has a similar effect:

https://developer.arm.com/documentation/102102/0103/What-is-Helium-#:~:text=Arm Helium technology is the M-Profile Vector Extension,Learning (ML) and Digital Signal Processing (DSP) applications.

ML is a subset of Artificial Intelligence (AI) that provides systems the ability to automatically learn and improve from experience without being explicitly programmed. Helium helps to boost Matrix Multiplication operations, which are the foundation of Convolutional Neural Networks or Classical based Machine Learning kernels.

Applications that can be greatly accelerated by Helium are Fast Fourier Transform (FFT) and Complex Dot Product as there are specific instructions which help implement these calculations.

Not sure if this has been posted before

Frontiers | Spiking Neural Network for Fourier Transform and Object Detection for Automotive Radar

The development of advanced autonomous driving applications is hindered by the complex temporal structure of sensory data, as well as by the limited computat...

www.frontiersin.org

So Infineon are checking out SNN's

Do we have more?