BRN Discussion Ongoing

[Diogenese]

“This is also the fastest, most energy efficient embodiment of Akida. Akida's 2-bit and 4-bte modes re more accurate, but are slower and use more power.”

While I am not technical at all my read of @Diogenese excellent post is that Akida and Plumerai are a great match for each other.

I'm on the fence here. There is a tech which is gaining some notoriety called in-memory-compute. I'm not sure how much this differs from Akida, which also has memory distributed with the compute elements.

What concerns me is that Plumerai are offering IP licences for FPGAs. Now, I suppose that, if they were to be using Akida IP, they could have a right to sub-licence Akida IP.

https://plumerai.com/#bnns

IP-core for BNN inference on FPGAs​

For customers that use FPGAs and require the most energy-efficient solution, we provide a custom IP-core that is highly-optimized for our BNN models and software.

The reason that 1-bit binarization does not lose too much accuracy when compared with 8, 16, or 32 bits is explained by Simon Thorpe's discussion of the JAST Rules.

Basically, the speed at which the optical nerve responds to light stimulus is proportional to the strength of the light stimulus. This means that, for a camera light sensor, the first pixel to respond is the most important, hence the winner-take-all rule.

This is Simon Thorpe's presentation: "The precision tradeoff" which explains temporal coding. It's the best explanation of "spikes good - floating point bad" that you could hope to find.

https://videos.insa-rennes.fr/video...6ac53873ac3153d6eae4a8812d22c9007c4f1bbcb1ed/

This slide illustrates the neuron receiving the stronger light signal (high intensity) firing before the neuron receiving the low intensity signal.

Rate code BAD!


This slide shows the intensity/delay function of the retina/pixel demonstrating that the stronger the light signal, the faster the response.

If you want to save your soul from hell,
a'ridin' on our range,
then cowboy change your ways today

and get this presentation between your ears.

 

[Deleted member 118]

So far this year we’ve had only a few companies confirm they are going to use Akida, but compare this to our suspicions it only a matter of time before things really start getting interesting.

 

[supersonic001]

Dan Lohrmann on LinkedIn: Google's DeepMind says it is close to achieving human-level AI | 79 comments

Google's DeepMind says it is close to achieving 'human-level' artificial intelligence – but it still needs to be scaled up --- Any thoughts on this?… | 79 comments on LinkedIn

www.linkedin.com

Hmmmm interesting, will we maintain a 3yr advantage or the juggernaut in pursuit

 

[Fact Finder]

Here are a few dots involving Nviso. -( @chapman89 - please note)

Nviso is on the Twitter page of xSoar promoting a podcast relating to Cortex XSoar automation security. When I visit Cortex XSoar's website they are very large in websecurity and have a very long list of customers and are currently hiring engineers and software developers able to work on Micro frontend architecture.

Below I have put up two links the first explains as captioned the 5 reasons to adopt Micro frontend architecture and the second is from AWS which runs a Micro frontend architecture model.

If you are going to run Micro frontend architecture models and break up your platform into manageable bits it, in my opinion, makes sense to move some cybersecurity along with processing to the frontend enter Nviso and Brainchip's AKIDA.

Anyway I expect some amongst the 1,000 Eyes will be able to put more meat on these bones.

The interesting thing to me was that I did not find Nviso promoting this event except on CortexXSoar's Twitter. It also does provide a further link to AWS remembering that Peter van der Made has spoken of AKIDA's ability to monitor key strokes to ensure the user is who they have logged in as which teamed with Nviso and its ability to profile humans all starts to make a lot of sense in a devolved workplace ie remote working environment connected via AWS.

My opinion only DYOR
FF

AKIDA BALLISTA

https://www.sitepoint.com/micro-frontend-architecture-benefits/

5 Reasons to adopt Micro frontend architecture

https://aws.amazon.com/blogs/architecture/micro-frontend-architectures-on-aws/

AWS and Micro frontend architecture

https://xsoar.pan.dev/

https://twitter.com/hashtag/córtex

Tweet
See new Tweets

Conversation

NVISO

@NVISOsecurity

Join us for a deep dive into #Cortex #XSOAR automations! In our upcoming webinar we will show you advanced use cases of #automation together with practical examples and code snippets. Register now: https://loom.ly/rZFbd1E

SOAR For Fun and Profit

Event details

25/05/2022, 19:00 - 25/05/2022, 20:00

Time zone: (UTC+10:00) Canberra, Melbourne, Sydney

Microsoft Teams meeting

The SOAR For Fun and Profit webcast will provide you a deep-dive into technical topics for Palo Alto Cortex XSOAR. We will share our knowledge we gained the last couple of years while developing automated workflows for the NVISO Managed Detect and Response Server.

In this first webcast episode we will do a deep dive into Cortex XSOAR automations where we will show you some advanced use cases of automation together with practical examples and code snippets.

Attendees will learn:

• Developing automations in Cortex XSOAR
• Using the Cortex XSOAR API in automations
• Executing commands in automations

Speakers

Wouter Stinkens

This was a rabbit hole in which I ended up lost so I asked Jesse @chapman the following question:

“Hi Jesse
I am having trouble with Nviso:

1. There is a Nviso that appears to have been started by IBM
2. There is Nviso Labs
3. There is Nviso Security
4. There is Nviso Birds
5. There is Nviso that partnered with BRN

They all appear to have started life in Europe.

Do you know if they are connected and where IBM currently fits in.”

Peter van der Made has secret sauce but Fact Finder has secret Jesse who is more than a match and it turns out Jesse has been able to confirm Nviso Security and hence Nviso Birds is a different company with the same name and in the same industry.

Further information will be provided as it emerges.

FF

AKIDA BALLISTA

 

[David holland]

Does anybody know Rob 's email, I'd like to send a nice email as I really like and rate this guy

 

[TheFunkMachine]

Ding dong, somebody say smart doorbell?

View attachment 7222

View attachment 7223

View attachment 7224

I know this sounds a lot like Akida. It cirtanly would be a good fit for their use cases. But are we jumping the gun and becoming to exited to quick here?

The thing I get caught up in is the fact they say the use deep learning and use something Called BNN (Binarized neural networks)

I know that Brainchip can make any deep learning algorithms into SNN with their SNN2CNN tool kit, but that makes every AI company a possible fit for Akida so I wouldn’t use that as an argument.

What am I missing here?

 

[Stable Genius]

I'm on the fence here. There is a tech which is gaining some notoriety called in-memory-compute. I'm not sure how much this differs from Akida, which also has memory distributed with the compute elements.

What concerns me is that Plumerai are offering IP licences for FPGAs. Now, I suppose that, if they were to be using Akida IP, they could have a right to sub-licence Akida IP.

https://plumerai.com/#bnns

IP-core for BNN inference on FPGAs​

For customers that use FPGAs and require the most energy-efficient solution, we provide a custom IP-core that is highly-optimized for our BNN models and software.

This is Simon Thorpe's presentation: "The precision tradeoff" which explains temporal coding. It's the best explanation of "spikes good - floating point bad" that you could hope to find.

https://videos.insa-rennes.fr/video...6ac53873ac3153d6eae4a8812d22c9007c4f1bbcb1ed/

This slide illustrates the neuron receiving the stronger light signal (high intensity) firing before the neuron receiving the low intensity signal.


View attachment 7262


View attachment 7263


Rate code BAD!


This slide shows the intensity/delay function of the retina/pixel demonstrating that the stronger the light signal, the faster the response.

View attachment 7264


If you want to save your soul from hell,
a'ridin' on our range,
then cowboy change your ways today

and get this presentation between your ears.

Thanks for that soundtrack Dio.

I served on “The Green Ghost” D39 so that instantly brought back memories from a different chapter in my life!

Cheers

 

[thelittleshort]

The thing I get caught up in is the fact they say the use deep learning and use something Called BNN (Binarized neural networks)

I know that Brainchip can make any deep learning algorithms into SNN with their SNN2CNN tool kit, but that makes every AI company a possible fit for Akida so I wouldn’t use that as an argument.

What am I missing here?

Hey @TheFunkMachine, the tech is beyond me but @Fact Finder responded to this directly a couple of pages back

 

[Fact Finder]

Dan Lohrmann on LinkedIn: Google's DeepMind says it is close to achieving human-level AI | 79 comments

Google's DeepMind says it is close to achieving 'human-level' artificial intelligence – but it still needs to be scaled up --- Any thoughts on this?… | 79 comments on LinkedIn

www.linkedin.com

Hmmmm interesting, will we maintain a 3yr advantage or the juggernaut in pursuit

Well it seems that they will if you read the article and not just the headline because in the article the following direct statements are attributed to De Freitas from Deep Mind:

"De Freitas comments came in response to an opinion piece published on The Next Web that said humans alive today won't ever achieve AGI.

De Freitas tweeted: 'It's all about scale now! The Game is Over! It's about making these models bigger, safer, compute efficient, faster...'

However, he admitted that humanity is still far from creating an AI that can pass the Turing test – a test of a machine's ability to exhibit intelligent behaviour equivalent to or indistinguishable from that of a human.' - https://lnkd.in/gZsKu2nj?trk=public_post_share-update_update-text This link appears in the link you posted.

It is always wise to go to the source document.

As for the 3 year lead that is in relation to Edge computing on a chip with the ability of AKIDA.

When it comes to Artificial General Intelligence Peter van der Made's latest comment on this was that Brainchip is 5 to 7 years ahead of the competition.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Fact Finder]

I'm on the fence here. There is a tech which is gaining some notoriety called in-memory-compute. I'm not sure how much this differs from Akida, which also has memory distributed with the compute elements.

What concerns me is that Plumerai are offering IP licences for FPGAs. Now, I suppose that, if they were to be using Akida IP, they could have a right to sub-licence Akida IP.

https://plumerai.com/#bnns

IP-core for BNN inference on FPGAs​

For customers that use FPGAs and require the most energy-efficient solution, we provide a custom IP-core that is highly-optimized for our BNN models and software.

This is Simon Thorpe's presentation: "The precision tradeoff" which explains temporal coding. It's the best explanation of "spikes good - floating point bad" that you could hope to find.

https://videos.insa-rennes.fr/video...6ac53873ac3153d6eae4a8812d22c9007c4f1bbcb1ed/

This slide illustrates the neuron receiving the stronger light signal (high intensity) firing before the neuron receiving the low intensity signal.


View attachment 7262


View attachment 7263


Rate code BAD!


This slide shows the intensity/delay function of the retina/pixel demonstrating that the stronger the light signal, the faster the response.

View attachment 7264


If you want to save your soul from hell,
a'ridin' on our range,
then cowboy change your ways today

and get this presentation between your ears.

MegaChips is offering AKIDA IP licences to customers if they want to add Ai to their existing chips if this allows you to be more comfortable.

My opinion only DYOR (and mine too please)
FF

AKIDA BALLISTA

 

[Bravo]

Great work TLS! 1000eyes strike again. Made my morning. Have a great weekend all.

Plumerai

Ultra-efficient and accurate AI solutions for Smart Home Cameras and IOT. Support for <br/>Cortex-M & A, Ethos-U. Fastest inference engine in the world.

www.arm.com

Plumerai looks to be partnered with ARM

Brainchip Plumerai ARM

Ooooohhhh! Texas Instruments and Plumerai!

 

[Bravo]

Great work TLS! 1000eyes strike again. Made my morning. Have a great weekend all.

Plumerai

Ultra-efficient and accurate AI solutions for Smart Home Cameras and IOT. Support for <br/>Cortex-M & A, Ethos-U. Fastest inference engine in the world.

www.arm.com

Plumerai looks to be partnered with ARM

Brainchip Plumerai ARM

Plumerai using very similar language to Mercedes.

PLUMERAI

MERCEDES

BRAINCHIP

 

[Dhm]

Here's Tesla's Q3 2020 earnings report presented by Elon Musk, which was referred to in the above article. I've also attached a patent "System and Method for Adapting a Neural Network Moodel on a Hardware Platform". Looks like Tesla could adopt Akida if it wanted to.


At 25.30 mins Elon describes the neural network approach and working with no cell phone connectivity.

WIPO - Search International and National Patent Collections

If the forecast that Elon was to at least explore or even better seriously consider us as a viable option of delivering his call of no cloud connection then that would lift our SP more than Mercedes did. If he actually chose us then fly me to the moon!

 

[Diogenese]

Thanks for that soundtrack Dio.

I served on “The Green Ghost” D39 so that instantly brought back memories from a different chapter in my life!

Cheers

She has a proud record:

https://www.navy.gov.au/hmas-hobart-ii

Hobart was awarded the coveted Duke of Gloucester’s Cup for being assessed as the most efficient unit of the RAN fleet. She was to win this cup on seven further occasions (1970, 1975, 1979, 1982, 1988, 1992 and 1994).

In recognition of the professionalism of her crew during her first deployment to Vietnam she was awarded a US Navy Unit Commendation for ‘exceptionally meritorious service’ as an element of the US Seventh Fleet during the period 10 March to 20 September 1967. The award was presented to Captain Griffiths on 21 April 1969. The citation highlighted the “outstanding teamwork, courage and professionalism displayed by Hobart’s officers and men.”
...
On 3 January 1975 Hobart, in company with Supply and Vendetta entered Darwin Harbour and the first working parties were landed. Members of the ship’s company undertook clearing and rehabilitation tasks in support of Operation NAVY HELP DARWIN. The ship quickly settled into a routine: over 100 men were landed daily and they worked two days ashore and had one day on board.

 

[Diogenese]

MegaChips is offering AKIDA IP licences to customers if they want to add Ai to their existing chips if this allows you to be more comfortable.

My opinion only DYOR (and mine too please)
FF

AKIDA BALLISTA

Thanks - I got distracted on the first half of that post - thank goodness it wasn't a barb wire fence.

 

[Dhm]

Hi Ray,

Great work! Can you twitt this as well? He-he-he!



View attachment 7248

View attachment 7250
View attachment 7251

I would be interested in just how much Akida has influenced Mercedes‘ battery efficiency. I don’t think that information is actually published but if it is it would be a brilliant calling card for us to use for all other EV manufacturers. Is anyone aware of this?

 

[Rise from the ashes]

Ooooohhhh! Texas Instruments and Plumerai!







View attachment 7266

 

[Stable Genius]

She has a proud record:

https://www.navy.gov.au/hmas-hobart-ii

Hobart was awarded the coveted Duke of Gloucester’s Cup for being assessed as the most efficient unit of the RAN fleet. She was to win this cup on seven further occasions (1970, 1975, 1979, 1982, 1988, 1992 and 1994).

In recognition of the professionalism of her crew during her first deployment to Vietnam she was awarded a US Navy Unit Commendation for ‘exceptionally meritorious service’ as an element of the US Seventh Fleet during the period 10 March to 20 September 1967. The award was presented to Captain Griffiths on 21 April 1969. The citation highlighted the “outstanding teamwork, courage and professionalism displayed by Hobart’s officers and men.”
...
On 3 January 1975 Hobart, in company with Supply and Vendetta entered Darwin Harbour and the first working parties were landed. Members of the ship’s company undertook clearing and rehabilitation tasks in support of Operation NAVY HELP DARWIN. The ship quickly settled into a routine: over 100 men were landed daily and they worked two days ashore and had one day on board.

History hey. I was onboard it in 92 when we won that award.

I’ve still got the actual Flag version of the Commendation from the US (not that I deserve that as I wasn’t there then) and commissioning pendant hanging in my games room!

Great work everyone with all the research. I’ve been flat out at work so apologies for not contributing much research lately!

Dio’s post just took me back in time!

Cheers

 

[Diogenese]

Renault morphoz:

Facial recognition and gesture control. It is an ev which means Akida?

Watch from 3:06

Not the right clip, but M. Hulot designed the extender car in the 60s:

https://www.bing.com/videos/search?...B04598171D6606D3F383B04598171D66&&FORM=VDRVRV

 

[Bravo]

Hey, ummmm...it just occurred to me that the 1000 's could be really sneaky and start looking for products that are being advertised with these parameters.