BRN Discussion Ongoing

Townyj

Ermahgerd
Wow and our man in Germany!

Strong automotive and battery background

View attachment 31948
1678610280491-png.31953


View attachment 31949
View attachment 31951
View attachment 31952

Bloody Hell... Some crazy Experience from all 3 of these. Nintendo, Analog Devices, Arm, Infineon, Broadcom.

Wow... My jaw dropped scrolling.
 
  • Like
  • Fire
  • Love
Reactions: 42 users
Agree, RF could be an interesting one.

Not just military but also Lidar amongst others.

I saw @Fact Finder earlier post discussing 16nm as well as Xilinx.

From Nandans tweet I decided to have a closer look at Ipsolon Research as he provided a web link. One of the companies he thanked re Akida Gen 2.

View attachment 31947

Was curious where they fit into picture.

So, snip of Ipsolon.


Ipsolon Research is an innovator in software defined radio technology.​


We launched Ipsolon Research in 2017 to develop and improve the analog and digital signal processing system for an airborne obstruction penetrating LIDAR system. Since then, we’ve engineered hardware, software, and FPGA design for a range of Software Defined Radios (SDR) and advanced wireless systems.

In addition to work for commercial wireless IC, we also manufacture and sell Chameleon and Cerberus Radios, both are wideband, small-form factor SDR for military communications, anti-jamming, and other advanced wireless signal processing applications. Our goal is to provide the wireless community with advanced radio system software and hardware development, early SDR conceptual design, advanced signal processing, and rapid prototyping.

From Ipsolon product page...16nm Xilinx.

They have a new product due soon but doesn't give specs yet.

ARAGMA​

Coming soon!


DIGITAL SIGNAL PROCESSING HORSEPOWER

16 TX/RX Channels, Smallest Footprint Software Defined Radio/RADAR

Multi-processor Subsystem

  • Xilinx 16nm FinFET RFSoC XCZU29
  • Quad-core 1.2 GHz ARM Cortex-A53 64b with L1/L2, MMU, DMA
  • Dual 500MHz R5 real-time processor, power manager & secure boot
  • 4GB DDR4, Micro SD memory
  • Range of IO including UARTs, 1G Ethernet, USB3.0
  • Massive parallel AXI IO between PS and FPGA fabric
FPGA Resources

  • 30+ system logic cells (K), 16 32g SERDES
  • 4,272 DSP slices, 2.13 TMACs @ 500 Mhz
Ipsolon also connected with US Navy via SBIR in 2021.


Ipsolon Research wins follow-on contract with Navy​

September 9, 2021

Fredrick, MD – The United States Navy has selected Ipsolon Research’s SBIR for a phase II follow-on contract after a successful phase I demonstration of its Deep Learning (DL) techniques.

“We used DL techniques to replace or extend the PDW-based approach to radar signal detection, “explained Ipsolon Research CEO, John Shanton.

Deep learning (DL) algorithms are a form of Machine Learning (ML) that use neural network (NN) architectures to process data and predict the presence of data patterns often called prediction or inference.

The specific goal of this project is to develop a deployable electronic sensor optimized to use ML techniques for real-time detection of radar signals and other signals of interest.

“The phase I results demonstrated that ML is a viable method for detection of a known radar signal type in power- and space-efficient FPGA devices,” Shanton said.


Was reading an article from 2020 about..

Artificial Intelligence in Radio Frequencies​



Excerpt.

Application of Machine Learning in Cognitive Radio​

The concept of CR was first proposed by Joseph Mitola [1] in order for mitigating the scarcity in limited radio spectrum by improving spectrum allocation efficiency by allowing unlicensed users (cognitive radio users) to identify and transmit over the frequency bands which are already assigned to the licensed users (primary users), but idle over specific time/space (spectrum holes). Spectrum hole identification is usually achieved by sensing the radio frequency (RF) environment through a process called spectrum sensing. It should be noted that the title of CR is not limited only to unlicensed wireless users and implied to any wireless user that adds the cognition capability along with reconfigurability to its system function.

So, where had I heard of "cognitive radio"?...oh that's right...from the Ph I and Ph II NASA solicitation with Intellisense using Akida.

So some prev posts on NECR and related SBIRs as refresh.



You have to love coincidences that keep mounting as eventually they provide proof beyond doubt.

My opinion only DYOR
FF

AKIDA BALLISTA
 
  • Like
  • Love
  • Fire
Reactions: 32 users

Steve10

Regular

Nintendo article from October 2021.

Patent Reveals Nintendo Is Working On Upscaling Technology

Just made public, the patent was first placed in March 2020


Patents are always a rather intriguing thing to track in the technology industry - sometimes they represent ideas and products that'll never see the light of day, but there are occasions when they give clues to future releases. A freshly released US patent application could be the latter, as it addresses technology that could allow Nintendo to boost visuals in its hardware.

Opened on 25th March 2020 and released publicly yesterday (30th September), the application is titled 'systems and methods for machine learned image conversion', and though the initial language can be a headscratcher it is essentially an idea similar to NVIDIA's DLSS. That's short for Deep Learning Super Sampling in NVIDIA's case, which works on some of its GPUs to upscale image resolution and quality in real time, while also being remarkably efficient and ensuring that the graphics card is under less strain. It's impressive technology, and it's been front and centre in a lot of talk about how Nintendo could produce a new Switch-style portable device that outputs higher resolution visuals while still working with low power output. Digital Foundry has explored this in detail.

What makes this application intriguing is that Nintendo is clearly exploring this internally - a named party on the application is Alexandre Delattre, who is a co-founder of Nintendo European Research and Development. It's also acknowledged in the 'Introduction' of the patent that this is an area being explored throughout the industry:

Machine learning can give computers the ability “learn” a specific task without expressly programming the computer for that task. One type of machine learning system is called convolutional neural networks (CNNs)—a class of deep learning neural networks. Such networks (and other forms of machine learning) can be used to, for example, help with automatically recognizing whether a cat is in a photograph. The learning takes places by using thousands or millions of photos to “train” the model to recognize when a cat is in a photograph. While this can be a powerful tool, the resulting processing of using a trained model (and training the model) can still be computationally expensive when deployed in a real-time environment.
Image up-conversion is a technique that allows for conversion of images produced in a first resolution (e.g., 540p resolution or 960×540 with 0.5 megapixels) to a higher resolution (e.g., 1080p resolution, 1920×1080, with 2.1 megapixels). This process can be used to show images of the first resolution on a higher resolution display. Thus, for example, a 540p image can be displayed on a 1080p television and (depending on the nature of the up-conversion process) may be shown with increased graphical fidelity as compared to if the 540p image were displayed directly with traditional (e.g., linear) upscaling on a 540 television. Different techniques for image up-conversion can present a trade off between speed (e.g., how long the process takes for converting a given image) and the quality of the up-converted image. For example, if a process for up-converting is performed in real-time (e.g., such as during a video game), then the image quality of the resulting up-converted image may suffer.
Accordingly, it will be appreciated that new and improved techniques, systems, and processes are continually sought after in these areas of technology.

Ultimately it shouldn't be a surprise that Nintendo is researching upscaling through machine learning, as it is likely to be a vital factor should the company opt to retain a Switch-style form factor while offering greater graphical fidelity in the future. Whether Nintendo will still utilise NVIDIA technology in future devices is also interesting; if it develops its own solution it may not need NVIDIA's DSSL tools. Of course, depending on what and who you believe, there are reports that '4K' development units are already out in the wild.

 
  • Like
  • Fire
  • Thinking
Reactions: 25 users

TECH

Regular
Now this has to be a significant large black dot made with permanent ink.

Great pick up @thelittleshort

My opinion only DYOR
FF

AKIDA BALLISTA

I totally second that !!...that is what I call excellent sleuthing...top stuff shortie ;)(y)

MegaChips-Nintendo-Brainchip (potentially)

Next week could be explosive at the Embedded World Conference, I just have this gut feeling, so pleased that Anil is going to be there
to represent us with all our key guys, I don't know the last time he has left the US, because he is such a workaholic, and I say that with the
upmost respect.

If the long termers think of how we started out, and I'm referring to presentations, interviews, roadshows, well, haven't we come a long
way, everything about our company reflects growth, not only staff numbers, I'm talking about growth as individuals, confidence in front
of a crowd, presentation slides, table cloths(?), website, diagrams...to say that we aren't growing is insane talk.

This could well be a major turning point for us, my friend arrives in Germany on location ready from the 14th, am I a little jealous, yes I am !
Getting to meet with Anil is a real blessing, something that I am looking forward to at some point, as I have mentioned before, he and Peter are
excellent founders/representatives of all that make Brainchip the great company that it is, honesty, work ethics, self discipline and good old fashioned kindness and respect.

Our booth is going to be oozing with a quiet confidence, surrounded by our partners within the same area, in knowing what we have, and where we are heading over the short term, just 1% of the pie, sorry, I have become greedy, how about 10% of the pie.

Stay tuned....the foundries are baking, not long now!!

Tech (Perth) :coffee:🙃
 
  • Like
  • Love
  • Fire
Reactions: 72 users

Kachoo

Regular
  • Like
  • Love
  • Fire
Reactions: 24 users

goodvibes

Regular

Intel Sorts out Supercomputing Future Amid Cancellation of GPUs​


Intel also has a set of additional AI accelerators with the AI training chips like Gaudi3, which is due this year, and its FPGAs. The chipmaker is also building incremental AI-acceleration units – mostly capable of inferencing – directly into its server chips. But GPUs will remain a mainstay for the company’s AI ambitions.
 
  • Like
  • Thinking
Reactions: 4 users

Diogenese

Top 20
Nintendo article from October 2021.

Patent Reveals Nintendo Is Working On Upscaling Technology

Just made public, the patent was first placed in March 2020


Patents are always a rather intriguing thing to track in the technology industry - sometimes they represent ideas and products that'll never see the light of day, but there are occasions when they give clues to future releases. A freshly released US patent application could be the latter, as it addresses technology that could allow Nintendo to boost visuals in its hardware.

Opened on 25th March 2020 and released publicly yesterday (30th September), the application is titled 'systems and methods for machine learned image conversion', and though the initial language can be a headscratcher it is essentially an idea similar to NVIDIA's DLSS. That's short for Deep Learning Super Sampling in NVIDIA's case, which works on some of its GPUs to upscale image resolution and quality in real time, while also being remarkably efficient and ensuring that the graphics card is under less strain. It's impressive technology, and it's been front and centre in a lot of talk about how Nintendo could produce a new Switch-style portable device that outputs higher resolution visuals while still working with low power output. Digital Foundry has explored this in detail.

What makes this application intriguing is that Nintendo is clearly exploring this internally - a named party on the application is Alexandre Delattre, who is a co-founder of Nintendo European Research and Development. It's also acknowledged in the 'Introduction' of the patent that this is an area being explored throughout the industry:






Ultimately it shouldn't be a surprise that Nintendo is researching upscaling through machine learning, as it is likely to be a vital factor should the company opt to retain a Switch-style form factor while offering greater graphical fidelity in the future. Whether Nintendo will still utilise NVIDIA technology in future devices is also interesting; if it develops its own solution it may not need NVIDIA's DSSL tools. Of course, depending on what and who you believe, there are reports that '4K' development units are already out in the wild.



US11494875B2 Systems and methods for machine learned image conversion

1678615717916.png


Although this patent does refer to NNs, I can't find any evidence that Nintendo have any in-house NPU design experience, so one is left to wonder where they could have got the NN from.
 
Last edited:
  • Like
  • Fire
  • Love
Reactions: 58 users

Quatrojos

Regular
I wonder if the ASX will be informed re the new reps in Japan and Germany...
 
  • Haha
  • Like
Reactions: 6 users

Labsy

Regular
GPT-3 used 10,000 Nvidia V100 GPU's !
ChatGPT - estimate is 4480 Nvidia A100 GPU's ( At $200,000 each, that's $896,000,000 )


I keep getting reminded of those old images of IBM'S 1st computer...massive, noisy and clunky....
I imagine a future of a complete true nuromorphic supercomputer, the size of a washing machine able to do what 10000 V100's can do with a fraction of the energy requirement... perhaps by 2030 as predicted by PVM...
 
  • Like
  • Love
  • Fire
Reactions: 11 users

equanimous

Norse clairvoyant shapeshifter goddess
Can anyone confirm you can order this in Psilocybin variant?

Screenshot_20230312_215042_Brave.jpg
 
  • Haha
  • Like
Reactions: 9 users

Diogenese

Top 20

Intel Sorts out Supercomputing Future Amid Cancellation of GPUs​


Intel also has a set of additional AI accelerators with the AI training chips like Gaudi3, which is due this year, and its FPGAs. The chipmaker is also building incremental AI-acceleration units – mostly capable of inferencing – directly into its server chips. But GPUs will remain a mainstay for the company’s AI ambitions.
'sfunny. I remember Donald Duck comics in which, when Scrooge used his computer to count his fortune, it used so much electricity that it caused a blackout.
 
  • Haha
  • Like
  • Love
Reactions: 25 users
Wow the 2 new additions appear to have excellent experience and connections.

Amazing bunch of people they are recruiting.

Thanks for the share.
Just a couple of Sunday night observations.

1. These are quality appointments.

Quality sales appointments have choices.

What attracts these types of employees?

Named accounts.

The advertisements stated they would be servicing named accounts.

2. They must be impressive named accounts in Japan & Germany.

My opinion only DYOR
FF

AKIDA BALLISTA
 
Last edited:
  • Like
  • Love
  • Fire
Reactions: 56 users

RobjHunt

Regular
Until Peter said that Akida could do maths, I did not believe that it could, and I still do not understand how it can.

Many years ago, philosophers and computer scientists used to talk about fuzzy logic. At the time, I did not follow the discussion:

https://en.wikipedia.org/wiki/Fuzzy_logic

von Neumann digital computers do not use fuzzy logic - it's a one or it's a zero.

Akida does something like "It's more likely to be one than zero", but not on an individual bit basis, rather on the basis of a batch of bits representing spikes, something I would classify as a form of fuzzy logic. Keep in mind that I am not familiar with what constitutes fuzzy logic.

I would think that things like N-of-M coding (rank coding) would be incompatible with mathematical logic.

There is much I do not understand about Akida.
Oh my lord DodgeyKnees, you're certainly not Robinson Crusoe ;)
 
  • Haha
  • Like
Reactions: 3 users

Easytiger

Regular
We've had a few changes there over the last few months since citi came on top of that list Deena.
Between July 2022 and October 2022 LDN's number came down, so did citicorp from 172 million to 151 million within 3 months.

And on the one you've shown it's 159 million. It's ever changing, and interesting how we have a "pool" of individual investors as our largest shareholder/s. Hopefully they're all friends 🤣
David James Evans ????
 

BaconLover

Founding Member
David James Evans ????
His holding hasn't changed in the last three years afaik.
 
  • Like
Reactions: 3 users

Diogenese

Top 20

Intel Sorts out Supercomputing Future Amid Cancellation of GPUs​


Intel also has a set of additional AI accelerators with the AI training chips like Gaudi3, which is due this year, and its FPGAs. The chipmaker is also building incremental AI-acceleration units – mostly capable of inferencing – directly into its server chips. But GPUs will remain a mainstay for the company’s AI ambitions.

Intel still think NNs use MACs (well at least they did in October 2022). Could they have learnt anything since?


US2023059976A1 DEEP NEURAL NETWORK (DNN) ACCELERATOR FACILITATING QUANTIZED INFERENCE US202218047415A·2022-10-18

1678618760190.png


PE = processing element.
[0008] FIG. 4 illustrates a PE (processing element) array, in accordance with various embodiments.

An DNN accelerator may include a PE (processing element) array performing MAC operations. The PE array may include PEs capable of MAC operations on quantized values. A PE may include subtractors for subtracting zeropoints from quantized activations and quantized weights to generate intermediate activations and intermediate weights. The intermediate activations and intermediate weights may be stored in data storage units in the PE and maybe used by an MAC unit in the PE. The subtractors may be placed outside the MAC unit but inside the PE. The MAC unit may perform sequential cycles of MAC operations. The MAC unit may include a plurality of multipliers. The intermediate activations and intermediate weights stored in the data storage units may be reused by different multipliers in different cycles of MAC operations. An output of the MAC unit or of the PE may be multiplied with a quantization scale to produce a floating-point value.


Intel has cancelled its Rialto Bridge processor, announced mid-2022, which was intended as an evolutionary change over its predecessor, and replaced it with Falcon Shores GPU which will be released in 2025, and its “flexible chiplet-based architecture will address the exponential growth of computing needs for HPC and AI,” McVeigh wrote.

... but does the left hand know which mouth the right foot is in?
 
  • Like
  • Haha
Reactions: 16 users

VictorG

Member
  • Haha
  • Like
Reactions: 6 users

Diogenese

Top 20
Yes you can.
Not sure about the car though.
It has mycoleather seats, but they taste like shitake to stop the passengers licking the seats.
 
  • Haha
  • Like
Reactions: 17 users

Kachoo

Regular
Just a couple of Sunday night observations.

1. These are quality appointments.

Quality sales appointments have choices.

What attracts these types of employees?

Named accounts.

The advertisements stated they would be servicing names accounts.

2. They must be impressive named accounts in Japan & Germany.

My opinion only DYOR
FF

AKIDA BALLISTA
This is not just for these two appointments either this goes for many of the latest additions to BRN.

I agree these people have left extreamly successful companies for some start up that has nothing absolutely no products customers or hope lol.

I would say it would take more time then it's worth but take a look at the caliber of personnel on any of the so-called competitors that BRN are being cased by and you will se the depth of the team.

I think that we would not have offices in

1. Germany if there was no money to be made
2. Japan cause there are economic leads for BRN,
3. INDIA,
4. France
5. Australia
6. USA
There could even be more. Looking at that that is pretty impressive actually. Lol

Cheers,
 
  • Like
  • Love
  • Fire
Reactions: 26 users
  • Haha
  • Like
Reactions: 20 users
Top Bottom