BRN Discussion Ongoing

[Fact Finder]

Here's a very interesting article published 5 hours ago. It's a Q&A with Michael Hurlston is CEO of Synaptics, a developer of hardware and software used in touchpads in computers, autos and smart home devices.. Lots of interesting info about their revenue, R&D spending, share price, difficulties in hiring, etc. TSMC are their largest supplier.

The article states "In March, Synaptics showcased innovations at tinyML Summit, including edge AI tech with low-power SoCs relying on neural network engines for vision, sound-detection and speech processing.

But what really stood out to me was this section...

Extract Only

FE: How does the demand for lower power in chips affect you?

Hurlston: With almost everything we do, power is an issue and so having lower power makes a difference. With AI at the edge, we know that a big driver of the battery in a phone is the display, so we have a face-detect AI algorithm. So, when the phone is close to the face when someone’s in the act of talking on the phone, we shut down the display. It’s a simple AI algorithm and we lock it in. Our phone manufacturing customers beat it up to death in testing, so they are not having fake shutdowns. They try to trick it a hundred different ways.

Here's another example of AI and machine learning… Today, we can do simple things like read license plates or read meters or count people. We can count people coming of a room. Instead of passing that task to a data center, you resolve it on a chip and the advantage is power savings where you don’t need to have a big engine, with passing back to the data center and the latency involved and ultimately the cost.

Take the example of a general purpose, low-power camera handed to a customer with tinyML to do something like identify sick chickens in a chicken coop that are sneezing. That’s a very specific use case, but a general purpose approach takes data and generates an ML model compiled onto a chip.

Q&A: Synaptics CEO Michael Hurlston

Michael Hurlston is CEO of Synaptics, a publicly-traded developer of hardware and software used in many products, including touchpads in computers, autos and smart home devices. | Synaptics innovates IoT as its stock plummets amid supply and labor shortages

www.fierceelectronics.com

Recently some thought incentivising Anil Mankar with options might not be what Brainchip should be doing.

I supported the move as I earnestly believe everything Brainchip does is extremely tactical. The background to why this was actually essential to ensure Brainchip’s market lead is disclosed in the following answer from Michael Hurston:

“Most grads want computer science jobs to work at places like Facebook and very few coming out are semiconductor engineers. The U.S. is losing competitiveness in this area. China is cranking out a lot of young engineers and they realize how important semiconductor engineering is. I feel that’s a major concern for anybody in this industry.

We have a footprint in China and a development center. There are super talented engineers in China, and of course there’s a big pull from state-owned enterprises and companies fueled by the government and they pay outrageous salaries. The multinational companies used to be the big payers, but that’s not true anymore.”

Anil Mankar’s skills in chip design are virtually irreplaceable and unique in the industry. Creating an environment where he will want to stay working and not retire is essential.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Fox151]

I see someone is familiar with Bluey terminology. Dollar bucks is the currency I pay for my make believe food with. A portion of 1% of a zillion billion dollar bucks would see me never going hungry again.

But trivialities aside, it truly is astounding how so many people today do seem to have the attention span of a gold fish and rely exclusively on technology to help them function.

It disgusts me that children have died from being left in cars because the parent “forgot” they were in there. It’s never a valid excuse.

As an ex-SES volunteer I have to admit smashing a window to get kids out was always our preferred method of rescue. We thought that might influence the parents to think twice next time.

 

[TasTroy77]

• Home
• Innovation and Technologies
• Technologies
• Interior cocoon

INTERIOR COCOON: ONBOARD TECHNOLOGIES FOR IMPROVED PASSENGER SAFETY​

• 
  

Valeo’s cocoon interior brings together several onboard technologies for improved passenger safety in the vehicle:

• a driver monitoring system, which uses a camera to analyze the driver’s face using artificial intelligence (AI) algorithms,
• a cabin monitoring system to detect occupant behavior, and
• the life detection system, which signals if a child or animal is in the rear of the locked vehicle.

WHAT IS A DRIVER MONITORING SYSTEM?​

A driver-monitoring system is an advanced safety technology that uses a camera to monitor driver alertness and check their level of vigilance. It is an innovative concept for active safety that can help reduce accidents on the road.
Valeo’s Driver Monitoring System has many functions such as detection of distraction and drowsiness, driver identification and facial emotion recognition.
When it detects signs of sleepiness or distraction, the system transmits alerts to the driver to get the driver’s attention back on the task of driving. The system’s camera mounted on the dashboard also ensures that the driver has their eyes on the road.

Other applications are also possible, such as driver identification for personalized settings and adapted driving modes.
Valeo’s Driver Monitoring System is in mass production with Deep Learning algorithms, including a scalable ECU and cameras for driver identification, accurate head-and-eye tracking, and monitoring driver gaze for distraction or drowsiness.
The driver monitoring system will contribute 25% of the ADAS evaluation and occupant condition monitoring to EuroNCap. The improvement of this automotive safety technology will contribute to the reduction of accidents related to fatigue and distraction, which account for 54% of fatal accidents in Europe. Camera-based driver identification also provides an additional level of security compared to key or badge-based access systems.

Since the DMS is becoming mandatory for L2 and L2+ autonomous vehicles, the global market is exploding, from about 1 million units in 2020 to an estimated 22 million units in 2025.

INNOVATIVE AUTOMOTIVE MONITORING SYSTEM FOR PASSENGER SAFETY​

WHAT IS A CABIN MONITORING SYSTEM?​

A cabin-monitoring system uses a camera sensor to understand human behavior and detect any passenger movements.
Thanks to the Valeo Interior Monitoring System, the vehicle is able to adapt to the internal context by knowing the characteristics of occupants such as their posture, and then adapting elements such as the air temperature or the driving mode. In addition, in the event of an accident, the intensity and deployment timing of the airbags can be adjusted according to the position and size of each passenger.

The cabin monitoring system, which is equipped with a camera, also allows entertainment functions such as taking selfies to share experiences while traveling.

Valeo Gesture recognitionoffers a natural and intuitive way to interact with the vehicle for even safer driving. Based on machine-learning algorithms and an in-house compact 3D camera, the feature is embedded in Valeo’s dome module.

A RADAR TO DETECT PRESENCE OF PASSENGERS IN THE VEHICLE​

The Life Presence Detection System uses interior radar and AI algorithms to detect if there’s life or not in the car. Result: no more children or pets forgotten in the back of the car anymore thanks to this additional safety technology. Once the engine is turned off and the car is locked, if the car detects that a person or a pet is still inside the vehicle, the system activates an audible and visual alarm on a smartphone.
Valeo’s “Life Presence Detection” solution is based on a 60 GHz radar using a millimeter wave detection system to detect in real-time occupants’ body movements, as subtle as breathing-related chest movements, even under visual obstacles such as clothing, blankets, etc.

Such systems with passenger movement detection are usually integrated in the vehicle roof liner and must be able to detect or assess the presence of a child or a pet in a locked vehicle and either provide warnings for as long as necessary or intervene to mitigate the risk of hyperthermia.
In Europe, EuroNcap has included life detection in its roadmap to address the problem of children left unattended in a vehicle, which leads to heatstroke injury and death.
Since 2022, this in-car monitoring feature is considered a key safety criterion in EuroNcap’s evaluation

Like in the new Mercedes "Awake " video

 

[Filobeddo]

Dont know if there's been any link previously found between BrainChip & AstraZenca

But the Head of IT Europe AstraZeneca, has a healthy ongoing interest in Brainchip on Linkedin. Or just a very keen shareholder

 

[Diogenese]

Recently some thought incentivising Anil Mankar with options might not be what Brainchip should be doing.

I supported the move as I earnestly believe everything Brainchip does is extremely tactical. The background to why this was actually essential to ensure Brainchip’s market lead is disclosed in the following answer from Michael Hurston:

“Most grads want computer science jobs to work at places like Facebook and very few coming out are semiconductor engineers. The U.S. is losing competitiveness in this area. China is cranking out a lot of young engineers and they realize how important semiconductor engineering is. I feel that’s a major concern for anybody in this industry.

We have a footprint in China and a development center. There are super talented engineers in China, and of course there’s a big pull from state-owned enterprises and companies fueled by the government and they pay outrageous salaries. The multinational companies used to be the big payers, but that’s not true anymore.”

Anil Mankar’s skills in chip design are virtually irreplaceable and unique in the industry. Creating an environment where he will want to stay working and not retire is essential.

My opinion only DYOR
FF

AKIDA BALLISTA

"Well Anil, the money or the box?"

Software engineers are no less brilliant than hardware engineers (if a little more abstract), but they are constrained by the hardware. There is an evident lack of eye-teeth in the hardware engineers at BrainChip. I reckon there would be a mile-long queue of hardware engineers hoping to work with Anil ... and a 10 mile long queue of HR poachers hoping to lure one away.

 

[HopalongPetrovski]

Here's a very interesting article published 5 hours ago. It's a Q&A with Michael Hurlston is CEO of Synaptics, a developer of hardware and software used in touchpads in computers, autos and smart home devices.. Lots of interesting info about their revenue, R&D spending, share price, difficulties in hiring, etc. TSMC are their largest supplier.

The article states "In March, Synaptics showcased innovations at tinyML Summit, including edge AI tech with low-power SoCs relying on neural network engines for vision, sound-detection and speech processing.

But what really stood out to me was this section...

Extract Only

FE: How does the demand for lower power in chips affect you?

Hurlston: With almost everything we do, power is an issue and so having lower power makes a difference. With AI at the edge, we know that a big driver of the battery in a phone is the display, so we have a face-detect AI algorithm. So, when the phone is close to the face when someone’s in the act of talking on the phone, we shut down the display. It’s a simple AI algorithm and we lock it in. Our phone manufacturing customers beat it up to death in testing, so they are not having fake shutdowns. They try to trick it a hundred different ways.

Here's another example of AI and machine learning… Today, we can do simple things like read license plates or read meters or count people. We can count people coming of a room. Instead of passing that task to a data center, you resolve it on a chip and the advantage is power savings where you don’t need to have a big engine, with passing back to the data center and the latency involved and ultimately the cost.

Take the example of a general purpose, low-power camera handed to a customer with tinyML to do something like identify sick chickens in a chicken coop that are sneezing. That’s a very specific use case, but a general purpose approach takes data and generates an ML model compiled onto a chip.

Q&A: Synaptics CEO Michael Hurlston

Michael Hurlston is CEO of Synaptics, a publicly-traded developer of hardware and software used in many products, including touchpads in computers, autos and smart home devices. | Synaptics innovates IoT as its stock plummets amid supply and labor shortages

www.fierceelectronics.com

As in space, so too here on Earth.........

"Power is Everything"

 

[robsmark]

Quatrojos says WANCA is an appropriate acronym for those that utter the words “What's.A.Neuromorphic.Chip.Anyway“.

Perhaps Sunny is an acronym for Second-class Unbearable Nasty Nauseating Yobo

Personally, I think he’s a Completely Unbearable Nasty Tormentor.

 

[Esq.111]

When I read articles like this and consider how once upon a time in a world not that far away parents, grandparents, friends and strangers had the ability to remember for more than ten seconds that they had the care of a child and they were in the car with them and now Governments around the world are legislating to make car manufacturers include intelligent devices to supplement humans after their ten second memory span has been exhausted ALL I CAN SAY IS AS BLIND FREDDIE SAYS -

‘This market that AKIDA technology can service is growing like a house on fire and will be worth a zillion billion dollar bucks.

The only way Brainchip will not capture at least one percent of this market is if they get high on cocaine and stay in bed for the next three years.

They truly are disrupting markets that do not exist yet and which are still to be thought of or imposed by Governments.’

My opinion only DYOR
FF

AKIDA BALLISTA

Afternoon Fact Finder,

I thought that's why executives are refferd to as the " C - suite "....

But yes, on a serious note, if management can't capture 1% Of the addressable market something has gone astray.

Personally, something in the order of 6% to 17% of the addressable market would please me greatly.
And dose not seem out of order, with everything going on in the broad spectrum of industry engagements, THAT WE KNOW OF.

Great work all, with all the new dots & new finds, just staggering, and here I was thinking the Easter long break would be a quiet period.

Happy Easter all Chippers.

Regards,
Esq.

 

[Deleted member 118]

@Fact Finder unless I’ve posted it before

http://eehpc.csee.umbc.edu/publications/pdf/2021/2105.05956.pdf

Now just found the 2022 version so I guess you have

Radware Bot Manager Captcha

iopscience.iop.org

 

[Boab]

@Fact Finder unless I’ve posted it before

http://eehpc.csee.umbc.edu/publications/pdf/2021/2105.05956.pdf

More and more articles keep popping up about the importance of saving power.
From mining Crypto currencies to door bells, this is all playing out very nicely for us Chippers.
Go you good thing.

 

[Boab]

More and more articles keep popping up about the importance of saving power.
From mining Crypto currencies to door bells, this is all playing out very nicely for us Chippers.
Go you good thing.

Ps
probably just as important is the issue of overheating...less power, less heat.

 

[Deleted member 118]

Quantum Computing Memristor To Unlock AI | New AI Designs Hypersonic Missile Without Humans.

 

[Space Cadet]

This NaNose study involved testing 10,000 people. By May 2021 they had tested about 7,000 and the study was extended for 12 months to test the remaining 3,000. When the testing is complete then they have to compile the results and report them to the FDA then wait for the FDA to advise what they decide. Still a lot of work to do.

My opinion only DYOR
FF

AKIDA BALLISTA

Talking about NaNose this trial on lung cancer should be finished late this year. Still chugging along

https://clinicaltrials.gov/ct2/history/NCT01840150?A=12&B=13&C=Side-by-Side#StudyPageTop

SC

 

[Fact Finder]

Dont know if there's been any link previously found between BrainChip & AstraZenca

But the Head of IT Europe AstraZenca, has a healthy ongoing interest in Brainchip on Linkedin. Or just a very keen shareholder

Hi FB

I am not aware of any prior link but reading his description of what his job is and the diseases he is to target I am thinking NaNose Brainchip.

Astra Zeneca is rolling in cash like all the big pharma and looking for a place to spend it as some of their leading market drugs are due to come out of their exclusivity period.

What better investment would there be than a handheld diabetes, Covid, Cancer detection device/s.

NaNose will clearly have to partner with someone to commercialise the device. I had thought Siemens would have the inside running but nothing wrong with a bidding war.

The 1,000 Eyes might find that Professor Haick and OR Professor Marshall have an Astra Zeneca connection.

If it’s the later then this could bring Noisy Gut Belt and Biotome into the picture.

Certainly a worthwhile find and generously shared.

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Filobeddo]

Hi FB

I am not aware of any prior link but reading his description of what his job is and the diseases he is to target I am thinking NaNose Brainchip.

Astra Zeneca is rolling in cash like all the big pharma and looking for a place to spend it as some of their leading market drugs are due to come out of their exclusivity period.

What better investment would there be than a handheld diabetes, Covid, Cancer detection device/s.

NaNose will clearly have to partner with someone to commercialise the device. I had thought Siemens would have the inside running but nothing wrong with a bidding war.

The 1,000 Eyes might find that Professor Haick and OR Professor Marshall have an Astra Zeneca connection.

If it’s the later then this could bring Noisy Gut Belt and Biotome into the picture.

Certainly a worthwhile find and generously shared.

My opinion only DYOR
FF

AKIDA BALLISTA

Hi FF, yes Astra Z has big pockets and they appear to be targeting AI in disease detection and drug manufacturing as a minimum. Interesting that they‘ve had a relationship for example since 2020 with Qure.ai, an Indian startup in targeting tech for detection of lung cancer, which is obviously right in the biotome nanose wheel house.

Hopefully others with longer attention spans than I can come up with something more solid

 

[Deleted member 118]

Not sure how important this is and I can’t remember why I was searching Cryogenic, but I came across this and thought you might find these few pages worth reading @Fact Finder

https://aeneas-office.org/wp-content/uploads/2022/01/ECS-SRIA2022_vb.pdf

 

[Boab]

Not sure how important this is and I can’t remember why I was searching Cryogenic, but I came across this and thought you might find these few pages worth reading @Fact Finder

https://aeneas-office.org/wp-content/uploads/2022/01/ECS-SRIA2022_vb.pdf

Hi Rocket577,
maybe you can jump into a Cryogenic chamber with an instruction to be thawed out when the SP is @$10.
Should take out all the anxiety of SP fluctuations.

 

[Deleted member 118]

Hi Rocket577,
maybe you can jump into a Cryogenic chamber with an instruction to be thawed out when the SP is @$10.
Should take out all the anxiety of SP fluctuations.

 

[Fact Finder]

Not sure how important this is and I can’t remember why I was searching Cryogenic, but I came across this and thought you might find these few pages worth reading @Fact Finder

https://aeneas-office.org/wp-content/uploads/2022/01/ECS-SRIA2022_vb.pdf

You know this paper is 385 pages long and I went on to triple time and one half at 5pm you must have deep pockets. I am great value though as I have managed in the 18 minutes or so since you have put up this post to find the following most relevant to our present situation as holders of the world's most advanced technology stock. The paper overall is about what needs to be done not what has been done by whom.

Anyway the following proves the point as to why we are all invested here:

"2.1.5.3.3 For Embedded Intelligence

Training AI models can be very energy-demanding. As an example, according to a recent study101, the model training process for natural-language processing (NLP, that is, the sub-field of AI focused on teaching machines to handle human language) could end emitting as much carbon as five cars in their lifetimes102. However, if the inference of that trained model is executed billions of times (e.g., by billion users' smartphones), its carbon footprint could even offset the training one. Another analysis103, published by the OpenAI association, unveils a dangerous trend: "since 2012, the amount of compute used in the largest AI training runs has been increasing exponentially with a 3.5 month-doubling time (by comparison, Moore's law had a 2-years doubling period)". These studies reveal that the need for computing power (and associated power consumption) for training AI models is dramatically widening. Consequently, the AI training processes need to turn greener and more energy-efficient...........................



AI accelerators are crucial elements to improve efficiency and performances of existing systems (to the cost of more software complexity, as described in the next challenge, but one goal will be to automatize this process). For the training phase, the large amount of variable precision computations requires accelerators with efficient memory access and large multi-computer engine structures. In this phase, it is necessary to access large storage areas containing training instances. However, the inference phase requires low-power efficient implementation with closely interconnected computation and memory. In this phase, efficient communication between storage (i.e., the synapses for a neuromorphic architecture) and computing elements (the neurons for neuromorphic) are paramount to ensure good performances. Again, it will be essential to balance the need and the cost of the associated solution. For edge/power-efficient devices, perhaps not ultra-dense technologies are required; cost and power efficiency matter perhaps more than raw computational performances. Other architectures (neuromorphic) need to be further investigated and to find the sweet use case spot. One key element is the necessity to save the neuronal network state after the training phase as reinitializing after switch-off will increase the global consumption. The human brain never stops. It is also crucial to have a co-optimization of the software and hardware to explore more advanced trade-offs. Indeed, AI, and especially DL, require optimized hardware support for efficient realization.

New emerging computing paradigms such as mimicking the synapses, using unsupervised learning like STDP (Spike-timing dependent plasticity) might change the game by offering learning capabilities at relatively low hardware cost and without needing to access large databases.

Instead of being realized by ALU and digital operators, STDP can be realized by the physics of some materials, such as those used in Non-Volatile Memories. Developing solutions for AI at the edge (e.g., for self-driving vehicles, personal assistants, and robots) is more in line with European requirements (privacy, safety) and know-how (embedded systems). Solution at the extreme edge (small sensors, etc.) will require even more efficient computing systems because of their low cost and ultra-low power requirements................."

My opinion only DYOR
FF

AKIDA BALLISTA

 

[Deleted member 118]

You know this paper is 385 pages long and I went on to triple time and one half at 5pm you must have deep pockets. I am great value though as I have managed in the 18 minutes or so since you have put up this post to find the following most relevant to our present situation as holders of the world's most advanced technology stock. The paper overall is about what needs to be done not what has been done by whom.

Anyway the following proves the point as to why we are all invested here:

"2.1.5.3.3 For Embedded Intelligence

Training AI models can be very energy-demanding. As an example, according to a recent study101, the model training process for natural-language processing (NLP, that is, the sub-field of AI focused on teaching machines to handle human language) could end emitting as much carbon as five cars in their lifetimes102. However, if the inference of that trained model is executed billions of times (e.g., by billion users' smartphones), its carbon footprint could even offset the training one. Another analysis103, published by the OpenAI association, unveils a dangerous trend: "since 2012, the amount of compute used in the largest AI training runs has been increasing exponentially with a 3.5 month-doubling time (by comparison, Moore's law had a 2-years doubling period)". These studies reveal that the need for computing power (and associated power consumption) for training AI models is dramatically widening. Consequently, the AI training processes need to turn greener and more energy-efficient...........................



AI accelerators are crucial elements to improve efficiency and performances of existing systems (to the cost of more software complexity, as described in the next challenge, but one goal will be to automatize this process). For the training phase, the large amount of variable precision computations requires accelerators with efficient memory access and large multi-computer engine structures. In this phase, it is necessary to access large storage areas containing training instances. However, the inference phase requires low-power efficient implementation with closely interconnected computation and memory. In this phase, efficient communication between storage (i.e., the synapses for a neuromorphic architecture) and computing elements (the neurons for neuromorphic) are paramount to ensure good performances. Again, it will be essential to balance the need and the cost of the associated solution. For edge/power-efficient devices, perhaps not ultra-dense technologies are required; cost and power efficiency matter perhaps more than raw computational performances. Other architectures (neuromorphic) need to be further investigated and to find the sweet use case spot. One key element is the necessity to save the neuronal network state after the training phase as reinitializing after switch-off will increase the global consumption. The human brain never stops. It is also crucial to have a co-optimization of the software and hardware to explore more advanced trade-offs. Indeed, AI, and especially DL, require optimized hardware support for efficient realization.

New emerging computing paradigms such as mimicking the synapses, using unsupervised learning like STDP (Spike-timing dependent plasticity) might change the game by offering learning capabilities at relatively low hardware cost and without needing to access large databases.

Instead of being realized by ALU and digital operators, STDP can be realized by the physics of some materials, such as those used in Non-Volatile Memories. Developing solutions for AI at the edge (e.g., for self-driving vehicles, personal assistants, and robots) is more in line with European requirements (privacy, safety) and know-how (embedded systems). Solution at the extreme edge (small sensors, etc.) will require even more efficient computing systems because of their low cost and ultra-low power requirements................."

My opinion only DYOR
FF

AKIDA BALLISTA

Shit I’ve got away quite cheaply as I thought it had 544 pages so thats a bonus then lol