BRN Discussion Ongoing

[Quiltman]

A nice reflection on Prophesee success in 2022 by Luca ...

 

[Fact Finder]

Hi Fmf,

I think this bit is particularly interesting, especially the choice of example of a "hardware start-up with a novel way of processing neural networks".

To assist customers with such applications, SiFive developed the new Vector Coprocessor Interface Extension (VCIX, pronounced “Vee-Six”). VCIX allows for tight coupling between the customer’s SoC/accelerator and the X280. For example, consider a hardware AI startup with a novel way of processing neural networks or one that has designed a very large computational engine. Instead of designing a custom sequencer or control unit, they can simply use the X280 as a drop-in replacement. With VCIX, they are given direct connections to the X280. The interface includes direct access into the vector unit and memory units as well as the instruction stream, allowing an external circuit to utilize the vector pipeline as well as directly access the caches and vector register file.

View attachment 25865

VCIX is designed to interface the NN to the SiFive X280.

The VCIX is a high-performance direct-coupling interface to the X280 and its instruction stream. To that end, Asanovic noted that on the X280 with the new VCIX interface, the X280 is capable of sending 1,024 bits over onto the accelerator/external component each cycle and retrieving 512 bits per cycle, every cycle sustained over the VCIX interface. [Per cycle @ 300 MHz, for example].

On the other hand, Google seem reluctant to abandon their in-house MXU and TPU:

Cliff Young, Google TPU Architect, and MLPerf Co-Founder was also part of the SiFive announcement. As we’ve seen from other Google accelerators, their hardware team always looks to eliminate redundant work by utilizing off-the-shelf solutions if it doesn’t add any real value to design it themselves in-house.

For their own TPU accelerators, beyond the inter-chip interconnect and their highly-refined Matrix Multiply Unit (MXU) which utilizes a systolic array, much of everything else is rather generic and not particularly unique to their chip. Young noted that when they started 9 years ago, they essentially built much of this from scratch, saying “scalar and vector technologies are relatively well-understood. Krste is one of the pioneers in the vector computing areas and has built beautiful machines that way. But should Google duplicate what Krste has already been doing? Should we be reinventing the wheel along with the Matrix Multiply and the interconnect we already have? We’d be much happier if the answer was ‘no’. If we can focus on the stuff that we do great and we can also reuse a general-purpose processor with a general-purpose software stack and integrate that into our future accelerators.” Young added, “the promise of VCIX is to get our accelerators and our general-purpose cores closer together; not far apart across something like a PCIe interface with 1000s of cycles of delay but right next to each other with just a few 100s of cycles through the on-chip path and down to 10s of cycles through direct vector register access.”

Google are still dabbling with analog SNNs:
WO2020077215A1 TEMPORAL CODING IN LEAKY SPIKING NEURAL NETWORKS

View attachment 25866
Spiking neural networks that perform temporal encoding for phase-coherent neural computing are provided. In particular, according to an aspect of the present disclosure, a spiking neural network can include one or more spiking neurons that have an activation layer that uses a double exponential function to model a leaky input that an incoming neuron spike provides to a membrane potential of the spiking neuron*. The use of the double exponential function in the neuron's temporal transfer function creates a better defined maximum in time. This allows very clearly defined state transitions between "now" and the "future step" to happen without loss of phase coherence.

May be time for a little morphic resonance.

Footnote: Wonder if Cliff Young still has those gumboots?

“To that end, Asanovic noted that on the X280 with the new VCIX interface, the X280 is capable of sending 1,024 bits over onto the accelerator/external component each cycle and retrieving 512 bits per cycle, every cycle sustained over the VCIX interface. [Per cycle @ 300 MHz, for example].”

Now wasn’t it Edge Impulse who was comparing AKIDA running at 300 MHz with a GPU running at 900 MHz???

My opinion only DYOR
FF

AKIDA BALLISTA

 

[equanimous]

Would love to hear a podcast with Intel in the next couple of months.

 

[equanimous]

 

[Mt09]

Would love to hear a podcast with Intel in the next couple of months.
I’d love to hear Mike Davies talk about Akida

 

[stuart888]

Nice to see Brainchip and partners Nviso and Edge-AI are the top three results.

The no cloud needed to run inference combined with privacy are big pluses for the Akida implementation!

The ability to process AI on devices unlocks impressive benefits for companies.

1. Improved bandwidth efficiency
2. Reduced latency
3. Fewer size and weight constraints
4. Enhanced privacy
5. Lower hardware costs
6. No cloud, ability to function offline
7. High availability
8. Improved model accuracy
9. Sustainability advantages
10. Lower costs compared to cloud-based AI
11. Increased levels of automation

 

[Fact Finder]

Hey FF, you've often mentioned to have a plan when it comes to an investment in BRN. I'm wondering what this actually looks like (although I know the specific details differ for each individual).
Do you have a target SP in mind at which you'll reduce your holdings by a certain % to fund lifestyle, pay off any debt etc, while keeping the remainder of your portfolio intact until the next SP milestone? For example: when the SP reaches $5 reduce holdings x 30% (to pay off mortgage or buy a yacht) and then hold the remainder until, say $7.50 at which point reduce by another 10-15%.
If so, is there flexibility to adjust for fundamentals i.e. in the above example at $5 would you hold if an upcoming CES promised an announcement of new tech that could see the SP rise in the short term, hoping to skim off 30% at a higher price?

Not seeking investment advice but simply wondering if a plan involves 'targets' and how disciplined one should be if/when those targets are reached.

Of course this assumes that the SP will actually continue to rise

Cheers Fd

Hi @Foxdog

I could say “all of the above” because these are all things that could be in someone’s plan.

But where I will start is with the why bother to have a plan.

We all are hopefully learning everyday from our life experience and in my Police career I had drummed into me the importance of operational planning to deal with emergency situations.

The reasoning behind this is very simple it is harder to think and easier to make mistakes in situations of stress.

So the best time to think and make decisions is during periods of calm when you can stand back and in your own time weigh all your options.

This idea carried over into my work as a prosecutor and then as a lawyer.

As a lawyer I quickly realised that when clients came to me looking for a solution to whatever their problem was it was incredibly rare for the client to actually have thought through what the solution would look like and where they would go once it was achieved.

Risk reward was a foreign concept which never entered their thinking.

For example in a simple personal injury case the risk you might think is clear you could loose.

This however is often not the greatest risk for the client the greatest risk is winning but not being awarded sufficient having regard to what it cost to win and then receiving a conservative assessment of their loss by the court leaving them with insufficient to achieve their goal.

Thus it can often be the case that taking an early offer of settlement for what seems to be a modest amount can best satisfy their goal.

What I discovered however was that the majority of my clients were so stressed at and during settlement negotiations they could not think properly about the risk reward even when laid out in front of them in black and white at the time an offer was made and being discussed.

It was apparent that I needed to educate the client from the moment they walked into my office about the need to have a plan should an offer of settlement be made to them. It is mostly the case that settlement offers have a time limit on acceptance often as tight as until close of business on the day of offer.

I see this as entirely analogous to buying shares in the hope they will appreciate in value.

At any point an announcement like the Mercedes Benz reveal can come out of left field and things go crazy and you have a very limited amount of time to make decisions under severe stress.

Someone who had a plan as simple as I will sell all my holding when it hits $2.00 and pay off my mortgage early and I am happy to wait five years would have been greatly advantages by putting that plan into immediate effect.

The person without a plan would likely have missed the opportunity by the time they decided what to do the opportunity would have been lost.

In fact I am sure like me you have read posts by some here saying ‘I wish I had …. when the price hit $2.34.’

Plans do not need to be complicated or clever.

All they need to be is made before the event.

The clever part is having a plan and executing it.

Buying 50,000 shares for your child’s long term financial future is a plan. But how simple and uncomplicated is such a plan.

Having a target value that will net you an amount to pay off your mortgage 30 years ahead of schedule is a plan.

Again very simple but nonetheless paying off your mortgage is life changing for most people. I know it was for my wife and I.

My personal situation and plan is peculiar to me and has no application to anyone else so has no value.

I am strange in this regard and have lots of factors in play that definitely are peculiar to me.

Suffice to say my plan is to create generational wealth. The first part of my plan involves trading to average down to increase my overall share portfolio. The second part is to continuously control risk by researching and reading everything I can about the companies I hold. I am not a set and forget investor.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[stuart888]

Xailient is sure focused on privacy, right from their homepage image/text.

However, digging deeper, I think they still use the cloud a lot via what they call Data Masking. Their words: "Data masking protects sensitive information and refers to using pseudonyms, scrubbing, redaction, and de-identification measures to safeguard a person’s data".

Not sure if Xailient is friend/partner or competition. But there are potential partnership dots: Xailient - Sony - Nviso - Brainchip.

I do not know much about Xailient, but they do come up a lot in google searches.

Akida is so strong with Edge IoT ML no-cloud low-power. Brainchip can fine tune for each use-case, with privacy/less cloud top of mind of consumers. All good news to me.

 

[Fact Finder]

The following popped up and looks like something given the partners the 1,000 Eyes could keep in mind:

 

[Medina]

 

[Fact Finder]

The following popped up and looks like something given the partners the 1,000 Eyes could keep in mind:

View attachment 25886

This first article is tantalising and justifies further research on AT&T.

Consider the scenario where upon partnering with Ford they discover AKIDA and think we could use this for ???? In our communications network to ????.

So next thing they are sitting with CEO Sean Hehir and to his surprise say we are interested in seeing if AKIDA can ??? in our 5G communications network.

AT&T Labs | Our Work | Analytics and AI-based Automation

AT&T Labs | Our Work | Analytics and AI-based Automation

about.att.com

My speculation only so DYOR
FF

AKIDA BALLISTA

 

[equanimous]

This is the new Rave on Wheels. A party on the move and how Mercedes likes to Roll.

 

[stuart888]

Pretty cool when AI jobs are not just looking for Python and Tensorflow, but also Experience with Brainchip Akida! Perhaps this is more common than I realize. Edge-AI (our partner) is helping on this likely.

https://neurotechx.com/job/internship-position-in-ai-ml-and-metaoptics-part-time-full-time/

This job has Meta Optics, which is new to me. Look up: Meta optics, or metalenses, is a radical new flat lens technology that will disrupt optics used for sensing and imaging in consumer electronics, autonomous vehicles, AR/MR displays, and more.

 

[Fact Finder]

This first article is tantalising and justifies further research on AT&T.

Consider the scenario where upon partnering with Ford they discover AKIDA and think we could use this for ???? In our communications network to ????.

So next thing they are sitting with CEO Sean Hehir and to his surprise say we are interested in seeing if AKIDA can ??? in our 5G communications network.

AT&T Labs | Our Work | Analytics and AI-based Automation

AT&T Labs | Our Work | Analytics and AI-based Automation

about.att.com

My speculation only so DYOR
FF

AKIDA BALLISTA

This article from 2016 confirms the AT&T thinking and I would be confident that they would be very aware of the NASA work and published papers on cognitive communications utilising AKIDA in the ScAN program:

https://www.fiercewireless.com/tech/at-t-experimenting-artificial-intelligence-as-a-key-enabler-next-gen-wireless

Saving power and reducing memory and improving retrieval by storing meta data is also a common theme in their publications.

My speculation only DYOR
FF

AKIDA BALLISTA

 

[dippY22]

Hi @Foxdog

I could say “all of the above” because these are all things that could be in someone’s plan.

But where I will start is with the why bother to have a plan.

We all are hopefully learning everyday from our life experience and in my Police career I had drummed into me the importance of operational planning to deal with emergency situations.

The reasoning behind this is very simple it is harder to think and easier to make mistakes in situations of stress.

So the best time to think and make decisions is during periods of calm when you can stand back and in your own time weigh all your options.

This idea carried over into my work as a prosecutor and then as a lawyer.

As a lawyer I quickly realised that when clients came to me looking for a solution to whatever their problem was it was incredibly rare for the client to actually have thought through what the solution would look like and where they would go once it was achieved.

Risk reward was a foreign concept which never entered their thinking.

For example in a simple personal injury case the risk you might think is clear you could loose.

This however is often not the greatest risk for the client the greatest risk is winning but not being awarded sufficient having regard to what it cost to win and then receiving a conservative assessment of their loss by the court leaving them with insufficient to achieve their goal.

Thus it can often be the case that taking an early offer of settlement for what seems to be a modest amount can best satisfy their goal.

What I discovered however was that the majority of my clients were so stressed at and during settlement negotiations they could not think properly about the risk reward even when laid out in front of them in black and white at the time an offer was made and being discussed.

It was apparent that I needed to educate the client from the moment they walked into my office about the need to have a plan should an offer of settlement be made to them. It is mostly the case that settlement offers have a time limit on acceptance often as tight as until close of business on the day of offer.

I see this as entirely analogous to buying shares in the hope they will appreciate in value.

At any point an announcement like the Mercedes Benz reveal can come out of left field and things go crazy and you have a very limited amount of time to make decisions under severe stress.

Someone who had a plan as simple as I will sell all my holding when it hits $2.00 and pay off my mortgage early and I am happy to wait five years would have been greatly advantages by putting that plan into immediate effect.

The person without a plan would likely have missed the opportunity by the time they decided what to do the opportunity would have been lost.

In fact I am sure like me you have read posts by some here saying ‘I wish I had …. when the price hit $2.34.’

Plans do not need to be complicated or clever.

All they need to be is made before the event.

The clever part is having a plan and executing it.

Buying 50,000 shares for your child’s long term financial future is a plan. But how simple and uncomplicated is such a plan.

Having a target value that will net you an amount to pay off your mortgage 30 years ahead of schedule is a plan.

Again very simple but nonetheless paying off your mortgage is life changing for most people. I know it was for my wife and I.

My personal situation and plan is peculiar to me and has no application to anyone else so has no value.

I am strange in this regard and have lots of factors in play that definitely are peculiar to me.

Suffice to say my plan is to create generational wealth. The first part of my plan involves trading to average down to increase my overall share portfolio. The second part is to continuously control risk by researching and reading everything I can about the companies I hold. I am not a set and forget investor.

My opinion only DYOR
FF

AKIDA BALLISTA

Wonderful response and explanation.

 

[stuart888]

The Brainchip solutions to problems/use cases is massive. Kind of freaky big. Event based sensing gets Team-PVDM fired up!

 

[Bravo]

This looks a wee bit interesting...

NEWS DMED​

Mercedes-Benz Usa, Disney Advertising & Merkley+Partners Collaborate With The Famous Group For First-Ever College Football Mixed Reality Activation During Upcoming Telecast Of The College Football Playoff Semifinal At The Chick-Fil-A Peach Bowl​

December 29, 2022




Together, Mercedes-Benz USA, ESPN, The College Football Playoff, The Peach Bowl, Disney Advertising, and Merkley+Partners have collaborated with fan experience company, The Famous Group, to develop an unforgettable experience for fans during the College Football Playoff (CFP) Semifinal at the Chick-fil-A Peach Bowl on Dec. 31 (8 p.m. ET, ESPN).

As part of the luxury automaker’s broader investment in electric vehicle technology, the companies have teamed up to create a mixed reality activation, leveraging cutting-edge technology to race four of the iconic brand’s electric vehicles around Mercedes-Benz Stadium for a spectacle, both on-air and on-site. This collaboration marks the first time that mixed reality technology has been used on a live broadcast during a college football game and is the latest in a string of innovations for the luxury mobility brand.
“We at Mercedes-Benz USA, with the support of strategic AOR Merkley+Partners, are leading the global evolution of electric luxury,” said Monique Harrison, Head of Marketing, Mercedes-Benz USA. “Reintroducing this iconic brand through authentic brand experiences to a new generation of consumer that previously had little or no affinity with the automotive industry. Collaborations with music, film and technology brands like Disney are part of our journey to become the world’s most desirable electric luxury brand.”
In the CFP Semifinal at the Chick-fil-A Peach Bowl, the exhibition will take place right before kickoff of the second half of the game, directly following the Mercedes-Benz Halftime report. Featuring four different Mercedes-Benz electric vehicles, the 45-second segment will push interaction with the physical environment further than before with the goal posts, field and iconic in-stadium Halo video board.
“We are thrilled to showcase the collective creativity and innovation amongst all the teams we work with,” said Theresa Palmieri, VP, Sports Brand Solutions, Disney Advertising. “We are constantly looking for opportunities to reimagine the advertising experience for brands and we are especially excited to unveil such an innovative execution in collaboration with Mercedes-Benz during a marquee event during the college football postseason.”
Using first-of-its-kind mixed reality production techniques, The Famous Group will operate three tracked cameras for the activation, including two fixed cameras and a SkyCam for the duration of the segment.

“Our Mixed Reality team has created some of the most impactful activations for leagues, teams, and brands across the industry,” says Andrew Isaacson, Partner at The Famous Group. “We’re thrilled to help bring this innovative technology to a live college football broadcast for the first time. This campaign will showcase an iconic brand while also providing fans with a unique and immersive experience.”
This campaign comes on the heels of The Famous Group’s mixed reality activation with Gillette during ESPN’s Monday Night Football game between the NFL’s Chicago Bears and New England Patriots in October, which saw a 69 percent increase in brand recall compared to advertisements on competitive broadcasts.
Media Contacts:
Bukola Oyalabu | Bukola.Oyalabu@Disney.com
Kristen Smith | Kristen.T.Smith@Disney.com
Julian Galano (The Famous Group) | julian@22-spring.com

Mercedes-Benz Usa, Disney Advertising & Merkley+Partners Collaborate With The Famous Group For First-Ever College Football Mixed Reality Activation During Upcoming Telecast Of The College Football Playoff Semifinal At The Chick-Fil-A Peach Bowl”

Together, Mercedes-Benz USA, ESPN, The College Football Playoff, The Peach Bowl, Disney Advertising, and Merkley+Partners have collaborated with fan experience company, The Famous Group, to develop an unforgettable experience for fans during the College Football Playoff (CFP) Semifinal at the...

dmedmedia.disney.com

 

[Fact Finder]

Pretty cool when AI jobs are not just looking for Python and Tensorflow, but also Experience with Brainchip Akida! Perhaps this is more common than I realize. Edge-AI (our partner) is helping on this likely.

View attachment 25889
https://neurotechx.com/job/internship-position-in-ai-ml-and-metaoptics-part-time-full-time/

This job has Meta Optics, which is new to me. Look up: Meta optics, or metalenses, is a radical new flat lens technology that will disrupt optics used for sensing and imaging in consumer electronics, autonomous vehicles, AR/MR displays, and more.

Hi @stuart888

You are on a roll today so I thought to emphasis the importance of the above post I would throw in the following statement taken from a larger interview with Mike Davies contained in the enclosed link:

“So that being said, we’re not at the final version. This process (Intel 4) is still in development, so we aren't really seeing products. Loihi 2 is a research chip, so there's a different standard of quality and reliability and all these factors that go into releasing products. But it certainly means that the process is healthy enough that we can deploy chips and put them on subsystem boards, and remotely access them, measure their performance, and make them available for people to use. My team has been using these for quite some time, and now we're just flipping the switch and saying our external users can start to use them. But we have a ways to go, and we have more versions of Loihi 2 in the lab - it's an iterative process, and it continues even with this release.”

The key takeaways about Loihi 2

1. Loihi 2 is a research chip with a different standard of quality and reliability - (in other words not fit for commercialisation)

2. Loihi 2 has a ‘way to go’ and may not be the final iteration.

An Interview with Intel Lab’s Mike Davies: The Next Generation of Neuromorphic Research

www.anandtech.com

AKIDA IP commercialised in market and has been taped out by Renesas in an MCU.

Night and day is the difference between Loihi 2 and AKIDA 1.0

The next generation commercial AKIDA technology advance is so close we can taste it.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[TechGirl]

Hi @stuart888

You are on a roll today so I thought to emphasis the importance of the above post I would throw in the following statement taken from a larger interview with Mike Davies contained in the enclosed link:

“So that being said, we’re not at the final version. This process (Intel 4) is still in development, so we aren't really seeing products. Loihi 2 is a research chip, so there's a different standard of quality and reliability and all these factors that go into releasing products. But it certainly means that the process is healthy enough that we can deploy chips and put them on subsystem boards, and remotely access them, measure their performance, and make them available for people to use. My team has been using these for quite some time, and now we're just flipping the switch and saying our external users can start to use them. But we have a ways to go, and we have more versions of Loihi 2 in the lab - it's an iterative process, and it continues even with this release.”

The key takeaways about Loihi 2

1. Loihi 2 is a research chip with a different standard of quality and reliability - (in other words not fit for commercialisation)

2. Loihi 2 has a ‘way to go’ and may not be the final iteration.

An Interview with Intel Lab’s Mike Davies: The Next Generation of Neuromorphic Research

www.anandtech.com

AKIDA IP commercialised in market and has been taped out by Renesas in an MCU.

Night and day is the difference between Loihi 2 and AKIDA 1.0

The next generation commercial AKIDA technology advance is so close we can taste it.

My opinion only DYOR
FF

AKIDA BALLISTA

Morning FF,

I think the only reason they mention Loihi alongside Brainchip in the job ad is not many people out there have Brainchip experience due to us being so new, so if you have Loihi experience then great we will get you onboard to work with Akida

 

[Fact Finder]

Morning FF,

I think the only reason they mention Loihi alongside Brainchip in the job ad is not many people out there have Brainchip experience due to us being so new, so if you have Loihi experience then great we will get you onboard to work with Akida

I agree with you as always. Well at least so far.