BRN Discussion Ongoing

[Diogenese]

@Chilling - are you William H Macey? I've read your posts for a while and they remind me of the role he played in The Cooler. Friendly, slightly diffident but a real downer on everyone around him. It was a deliberate tactic by his casino employer to reduce the buzz of a successful punter. We're not punters, we're investors - and I have a suspicion you're not an investor like us - I think you may be a little escapee from that other place, just sowing seeds of doubt. If I'm wrong then tell me so and I'll back off while I read your positive posts about our great company. BAAAAAALLLLISTA

Loved his work in Fargo ... and the body count was only slightly less than the Korean War ...

... must dig something out of my Cohen Bros library tonight.

 

[Deleted member 118]

Another day as a Brainchip holder down a mere 8% never a dull moment cant turn your back for a second.
Really has me wondering on this 4C isnt it due next week?

Chin up as we are still up on the week.

 

[SmallVentures]

Not sure if this is the right channel for it, but I was hoping somebody could put into perspective Tesla’s Dojo D1 chip with its neural network accelerator, versus Akida.

A good friend of mine continuously emphasises how advanced Tesla’s D1 chip is for ADAS, and I try and explain to him that the fundamental architecture of Von Neumann will be outdated once Akida takes off.

Can anybody offer further input on how to best explain this and put it even more into perspective? Are there any specs between Dojo D1 and Akida 1000 that can be compared one to one?

 

[Fact Finder]

The CEO Sean Hehir clearly stated in a recent presentation where he was being interviewed by an uninspiring interviewer that income would commence to ramp up later this year. The interviewer repeated his answer to him and he gave an unqualified ‘YES’.

In his most recent interview/presentation he spoke to the fact that 2023 was when income and expenditure would balance each other.

My recollection of what is being said by the person claiming to be the CEO of Brainchip.

There is no valid reason or fact upon which to base a belief that this 4C will produce other than a modest increase in income over last quarter.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Mugen74]

Loved his work in Fargo ... and the body count was only slightly less than the Korean War ...

... must dig something out of my Cohen Bros library tonight.

Ot-My favorite

 

[Fact Finder]

Not sure if this is the right channel for it, but I was hoping somebody could put into perspective Tesla’s Dojo D1 chip with its neural network accelerator, versus Akida.

A good friend of mine continuously emphasises how advanced Tesla’s D1 chip is for ADAS, and I try and explain to him that the fundamental architecture of Von Neumann will be outdated once Akida takes off.

Can anybody offer further input on how to best explain this and put it even more into perspective? Are there any specs between Dojo D1 and Akida 1000 that can be compared one to one?

Not sure if this is a genuine question as the Dojo D1 uses 400 watts of electricity and is burning the planet to the ground. AKIDA uses milli watts of power and can help save the planet.

In autonomous vehicles Dojo D1 kills battery range whereas AKIDA according to Mercedes Benz helps to make over 1,000 kilometres per charge possible.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[equanimous]

Not sure if this is a genuine question as the Dojo D1 uses 400 watts of electricity and is burning the planet to the ground. AKIDA uses milli watts of power and can help save the planet.

In autonomous vehicles Dojo D1 kills battery range whereas AKIDA according to Mercedes Benz helps to make over 1,000 kilometres per charge possible.

My opinion only DYOR
FF

AKIDA BALLISTA

Smarter, faster and more efficient.

 

[HopalongPetrovski]

Loved his work in Fargo ... and the body count was only slightly less than the Korean War ...

... must dig something out of my Cohen Bros library tonight.

"Unbroken" might be apt, after todays performance.

 

[Diogenese]

Not sure if this is the right channel for it, but I was hoping somebody could put into perspective Tesla’s Dojo D1 chip with its neural network accelerator, versus Akida.

A good friend of mine continuously emphasises how advanced Tesla’s D1 chip is for ADAS, and I try and explain to him that the fundamental architecture of Von Neumann will be outdated once Akida takes off.

Can anybody offer further input on how to best explain this and put it even more into perspective? Are there any specs between Dojo D1 and Akida 1000 that can be compared one to one?

Hi SV,

Akida's forte is signal matching - image classification/identification, voice recognition, sensor output identification generally. von Neumann processors currently do this using CNN (convolutional neural network) software but they use enormous amounts of power performing the mathematical calculations involving 32 or 64 bits to do it.

As initially conceived, Akida using 1-bit weights and actuations does not use mathematical multiplications. Instead the single bits are considered as spikes. Akida does do addition of single bits to determine when the inputs to a neuron have reached firing threshold value, but this uses far less power than 32/64 bit multiplication. In addition, using image sensors as an example, in Akida a "spike" is only generated when the input from a pixel changes value, and the changes are treated asynchronously (as they occur, compared with frame-by-frame in a vN processor). In contrast, a von Neumann processor must perform a comparison for every 32/64 bit pixel on consecutive frames, using 1000 times the power that Akida uses to perform the same task.

In an AV, there is still a place for vN processors to use a program to act on the results of the Akida signal classification process, eg, by generating a brake signal or a turn signal for the AV, using the Akida output as a program parameter. However, Akida improves the overall power efficiency and speed of response by something of an order of magnitude or more because the CNN software utilizes so much power.

In simpler cases, the output from a stand-alone Akida may be sufficient to initiate a simple ON/OFF type function. However, in many instances, Akida will be run in conjunction with a CPU or MPU.

When higher accuracy is required, Akida can use 2-bit or 4-bit weights and actuations. I am just guessing that these will involve 2-bit or 4-bit multiplications, but 4-bit multiplication will still be 64 times (8*8) more energy efficient than 32-bit multiplication, as well as being 64 times faster. [That's 256 times better compared to 64-bit multiplications]. So Akida can trade off some power and speed for improved accuracy in performing these classification tasks, but it is still miles ahead of vN.

 

[Derby1990]

Ghost entry
You will never know my name
I could make Elon my cleaner
Just quietly buying up BRN shares
Then Akida in everything will allow me to buy the solar system
And only then will I expand.

 

[SmallVentures]

Not sure if this is a genuine question as the Dojo D1 uses 400 watts of electricity and is burning the planet to the ground. AKIDA uses milli watts of power and can help save the planet.

In autonomous vehicles Dojo D1 kills battery range whereas AKIDA according to Mercedes Benz helps to make over 1,000 kilometres per charge possible.

My opinion only DYOR
FF

AKIDA BALLISTA

It was a genuine question, but understand that’s it’s difficult to compare the two, therefore I was asking for perspectives …

Other than Power consumption / efficiency, which of course is a clear advantage, would it be fair to say that Unlike the D1, which still crunches data the traditional way and relies on Tesla’s supercomputer to keep improving its learning, Akida learns in real time, no latency, on chip without supercomputer?

 

[Diogenese]

Hi SV,

Akida's forte is signal matching - image classification/identification, voice recognition, sensor output identification generally. von Neumann processors currently do this using CNN (convolutional neural network) software but they use enormous amounts of power performing the mathematical calculations involving 32 or 64 bits to do it.

As initially conceived, Akida using 1-bit weights and actuations does not use mathematical multiplications. Instead the single bits are considered as spikes. Akida does do addition of single bits to determine when the inputs to a neuron have reached firing threshold value, but this uses far less power than 32/64 bit multiplication. In addition, using image sensors as an example, in Akida a "spike" is only generated when the input from a pixel changes value, and the changes are treated asynchronously (as they occur, compared with frame-by-frame in a vN processor). In contrast, a von Neumann processor must perform a comparison for every 32/64 bit pixel on consecutive frames, using 1000 times the power that Akida uses to perform the same task.

In an AV, there is still a place for vN processors to use a program to act on the results of the Akida signal classification process, eg, by generating a brake signal or a turn signal for the AV, using the Akida output as a program parameter. However, Akida improves the overall power efficiency and speed of response by something of an order of magnitude or more because the CNN software utilizes so much power.

In simpler cases, the output from a stand-alone Akida may be sufficient to initiate a simple ON/OFF type function. However, in many instances, Akida will be run in conjunction with a CPU or MPU.

When higher accuracy is required, Akida can use 2-bit or 4-bit weights and actuations. I am just guessing that these will involve 2-bit or 4-bit multiplications, but 4-bit multiplication will still be 64 times (8*8) more energy efficient than 32-bit multiplication, as well as being 64 times faster. [That's 256 times better compared to 64-bit multiplications]. So Akida can trade off some power and speed for improved accuracy in performing these classification tasks, but it is still miles ahead of vN.

I don't have the link to hand, but Simon Thorpe's presentation posted on here by @TECH a week or so ago explains why more than 4 bits are superfluous when he discusses N out of M coding. The trick being the recognition that stronger signals register on the optic nerve earlier than weaker signals. Thus only the first in time N signals are required to recognize the input.

As Simon says - this is the secret sauce.

 

[cosors]

"BrainChip and NVISO partner on human behavioural analytics in AI devices​

22 Apr 2022

BrainChip and nViso, a specialist in human behavioural analytics, have announced a collaboration targeting battery-powered applications in robotics and mobility/automotive to address the need for high levels of AI performance with ultra-low power technologies.

The initial effort will include implementing NVISO’s AI solutions for Social Robots and In-cabin Monitoring Systems on BrainChip’s Akida processors.
With developers of automotive and consumer technologies looking to develop devices that better respond to human behaviour, they require tools and applications that are capable of interpreting human behaviour that can be captured from cameras and sensors on devices.
These environments can be constrained, however, by limited compute performance, power consumption, and cloud connectivity lapses. Akida looks to address these weaknesses with high performance and ultra-low power (micro- to milliwatts) as well as by performing AI/ML processing of vision/image, motion, and sound data directly on devices, instead of in the cloud and since information is not sent off-device, user privacy and security are also protected.
NVISO’s technology analyses signals of human behaviour such as facial expressions, emotions, identity, head poses, gaze, gestures, activities, and objects with which users interact. In robotics and in-vehicle applications, human behaviour analytics detect the user’s emotional state to provide personalised, adaptive, interactive, safe devices and systems.
The result of the collaboration between NVISO and BrainChip is expected to enable more advanced, more capable, and more accurate AI on consumer products.
“Our work with BrainChip will support AI’s demanding power/cost/performance needs for OEMs, even at mass production and scale, so they can benefit from faster and more efficient development cycles,” said Tim Llewellynn, CEO of NVISO. “Ultra-low power edge-based consumer processing is expected to deliver a more intelligent and individualized user experience, and we believe running our AI solutions for Social Robots and In-cabin Monitoring Systems on Akida will provide a competitive edge for joint customers demanding always-on features on low power budgets.”
“NVISO’s human behavioural analytics AI systems offer fascinating possibilities in homes, cars, buildings, hospitals, and more, and we’re enthusiastic about supporting these functions with BrainChip’s processing performance and energy efficiency,” said Sean Hehir, BrainChip CEO. “This is not only a collaboration between two companies, it’s advancing the state of the art in AI with platforms for edge AI devices to interpret human behaviour, improving product performance and user experience.”
BrainChip’s neuromorphic processor, Akida, is able to mimic the human brain to analyse only essential sensor inputs at the point of acquisition, processing data with unparalleled efficiency, precision, and economy of energy. Keeping AI/ML local to the chip, independent of the cloud, also dramatically reduces latency while improving privacy and data security."
https://www.newelectronics.co.uk/co...-on-human-behavioural-analytics-in-ai-devices

 

[Labsy]

Now there is a lie for them to run with universal income. They could claim to be paying the dole to 300 million low income earners on distant planets around the universe how would we ever audit it. LOL

The naivety of the Australian electorate is staggering on the Solomon Island China issue. In the Wheat Board scandal the defence was the only way we can compete when selling to the tin pot regimes that pass themselves off as democratically elected is to pay bribes just like China and others in this space.

Australia was hauled over the coals for the payment of bribes to sell Australian wheat for Australia and made scape goats of those at the Board. The fantasists here in Australia joined with the United Nations in decrying this conduct with nations like China and Russia cheering them on as they just continued on their merry way paying bribes and assassinating people who disagreed with their democratically elected leaders.

Now ignoring the past and all the evidence it is Australia's fault that China has been able to hijack the Solomon Islands. The worlds former greatest Prime Minister Jacinta Ardern who was lauded by the left and middle of politics here in Australia with the way she has embraced her island brothers and sisters in the Pacific was also blindsided on this yet it is all Australia's fault.

We here need to accept that outside Australia the majority of the World's population live in countries where they do not play by the rules. They do not prosecute their own soldiers for killing the wrong people in fact they give them medals. They do not play fair in trade and bribery of corrupt officials is only criticised if they miss the deal because they did not offer enough to secure the required agreement or sale:

"The AWB oil-for-wheat scandal (also known just as the AWB scandal) refers to the payment of kickbacks to the regime of Saddam Hussein in contravention of the United Nations Oil-for-Food Humanitarian Programme. AWB Limited is a major grain marketing organisation based in Australia. For much of the 20th and early 21st century, it was an Australian Government entity operating a single desk regime over Australian wheat, meaning it alone could export Australian wheat, which it paid a single price for. In the mid-2000s, it was found to have been, through middlemen, paying kickbacks to the regime of Saddam Hussein, in exchange for lucrative wheat contracts. This was in direct contradiction of United Nations Sanctions, and of Australian law.

AWB delivered 90% of the Iraqi wheat market, before its practices were questioned in 2005. United Nations investigator Paul Volcker found that the Australian Wheat Board, and later AWB Limited, were not the only, but certainly the largest source of kickbacks to the Iraqi regime. The Australian Government also launched a Royal Commission, which recommended that criminal proceedings commence against 12 people. Ultimately, criminal charges were dropped by the Australian Federal Police. Several Australian civil cases were however successful. Since the payments were discovered, AWB Limited has undergone a major restructuring, losing its monopoly supply of Australia wheat exports, and appointing an entirely new management. However, its profitability continues to suffer.

Although AWB and by extension the Australian Government were not the only entities to be implicated in the Oil-for-Food scandal, the event earned a place in Australian political consciousness"

Our share market is allowed to be the play thing of overseas institutions on the spurious basis that we need liquidity and competition what an absolute joke.

Then these foreign manipulators transfer the profits offshore, do not pay tax, then bring back their ill gotten gains and buy our assets because it is great to have a safe harbour outside their own countries with the rule of law and a stable democracy that will protect their property rights and not imprison or prosecute them without due and fair process which requires evidence even though the whole world knows what they are doing.

The purpose of the above rant is to ensure ordinary retail investors understand that they cannot compete all they can do is try and get on the life boat before it is launched and stay seated even when while being lowered it starts to tilt because someone on the ship has stuck a spike in one of the rollers to try and make room on the life boat by tipping out and drowning a few retail holders who panic and let go of their seat.

My opinion only DYOR
FF

AKIDA BALLISTA

Well said FF... bloody brilliant

 

[Fullmoonfever]

Further to @Fact Finder s post about issues in the current climate attracting quality, experienced staff I see KBR advertised again through March (closed end Mar) for a RadNeuro Research Engineer...unless of course they need an additional one

Is the same details as the role advertised in Jan this year I posted HC in early Feb.

Good to see still working towards prototyping for NASA at least.

Latest first then HC snip.

RadNeuro Research Engineer job in Mountain View at KBR | Lensa

KBR is currently looking for a RadNeuro Research Engineer near Mountain View. Full job description and instant apply on Lensa.

lensa.com

 

[Bravo]

I feel a bit embarrassed posting this after the previous posts which have all been so eloquent, thoughtful and just plain bloody brilliant! But I just wanted to let everyone know that my horoscope prediction for today was spot on in the finance section, not once but TWICE!

In spite of how frivolous it may seem, it strikes me that it bodes EXTREMELY well for all of us!

Expenses = More BrainChip shares
Investment Gains: More BrainChip shares

 

[Labsy]

Accumulated more earlier this week and more today... was strangely hoping sp would drop down into the 93c mark. Why? Because, as many have repeated, we most likely won't see these prices again. Insto's definitely driving prices down, so why not use that to our advantage also? It's a long game. Revenue will increase this year but I fear we are investing too much emotion on these short term figures... 2025 is where the action will take place.
I also can't help but feel a little nervous about all the Mercedes-Benz hype. I think this will slowly pan out but I'm investing on the expected returns from the renesas and megachips deal... our IP will be in so many products, this is where the saturation will take place.
Every dollar (or 93 cents) invested today in BRN is 5 dollars in 2025 or slap me in the face.... If u prefer keep your cash under the mattress or take a holiday but I can wait for my vacation in 2025.

AKIDA BALLISTA!!!!

 

[Sirod69]

AI Summary - Trending News About Artificial Intelligence

Trending news about Artificial Intelligence

www.ai-summary.com

 

[Bravo]

I feel a bit embarrassed posting this after the previous posts which have all been so eloquent, thoughtful and just plain bloody brilliant! But I just wanted to let everyone know that my horoscope prediction for today was spot on in the finance section, not once but TWICE!

In spite of how frivolous it may seem, it strikes me that it bodes EXTREMELY well for all of us!

Expenses = More BrainChip shares
Investment Gains: More BrainChip shares


View attachment 4811

But then it went on to say this, which I am a bit weirded out by. But that's ok, I'm going to have a chamomile tea and clean my chakras out, so everything should be fine by tomorrow.

 

[Fact Finder]

Brainchip - AI Cutting Edge Chip - Brain Functions On A Chip - Rob Telson - VP Sales Brainchip

Topics: The future of AI within everyday technology like phones, smart homes, and transportation. The science behind ‘mimicking the brain’ – allowing art…

www.techprounicorn.com

Another podcast from Rob Telson

Haven't had a chance to listen myself yet but thanks for the find @Dallas

As poor old Molly used to say DO YOURSELF A FAVOUR and take a listen to this podcast.

I apologise in advance for the Vice President of World Wide Sales at Brainchip being over the top excited about the product he is selling called AKIDA Technology.

I also apologise for the podcast host being more excited about AKIDA Technology than Rob Telson and for his over the top statements at the end of the podcast that AKIDA technology is going to change the World.

I know some retail investors will get carried away and buy shares in Brainchip just because of their combined enthusiasm so please don’t do that actually listen to why they are so excited then decide whether to invest.

My opinion only DYOR
FF

AKIDA BALLISTA