BRN Discussion Ongoing

[manny100]

BrainChip ASX: Akida neuromorphic processor. AI at the edge. - B2B Marketing Agency
" 40 years ago, when Terminator 1 was released, a neural network based CPU was pure futuristic fantasy."
" It seems we have arrived at the future. Australian company Brainchip Holdings is about to launch the first of a series of chips supporting brain-like neural processing that will likely change the world."
Good read.
At the end of this lengthy, interesting and easy to read article there are 62 links to further reading.

 

[manny100]

Finding Bacteria in Blood: A Hardware-based Neuromorphic Solution for Real-World E-Nose Applications

Explore the revolutionary application of neuromorphic technology in creating an electronic nose for real-world scenarios.

brainchip.com

" The performance of this system is notable—when tested on the MedNose dataset, it demonstrated an impressive ability to distinguish among ten different bacteria types, achieving a peak classification accuracy of 97.42%. This not only surpasses previous models, which needed multiple samples to reach slightly lower accuracy levels, but also does so with remarkable efficiency. Critically, the SNN model was rigorously tested on the Akida NSoC, where it exhibited exceptionally low power usage, consuming only 24.5 mW, while maintaining a high throughput of 181 inferences per second. This performance marks a significant advancement, indicating that such a system is not only possible but now can be provided with a very cost-effective and portable device that enables rapid disease diagnosis with the critical level of precision."

 

[Rach2512]

The Power of Neuromorphic Computing: | Thomas Hülsing

The Power of Neuromorphic Computing: A Brain-Inspired Revolution Have you ever wondered how powerful and energy-efficient our brains are? While a human brain uses only about 20 watts of energy (less than a light bulb), it can perform complex tasks like recognizing faces, learning...

www.linkedin.com

 

[Space Cadet]

Speaking of sanctions:
I wonder what’s the punishment for violating the spelling of our company’s name ON OUR OWN WEBSITE?

View attachment 85548

Always look on the bright side of life.

I'm Brian and so is my wife.

SC

 

[Pom down under]

Speaking of sanctions:
I wonder what’s the punishment for violating the spelling of our company’s name ON OUR OWN WEBSITE?

View attachment 85548

I can’t see anything wrong

 

[Baron Von Ricta]

Always look on the bright side of life.

I'm Brian and so is my wife.

SC

Hey SC,
Life of Brian grossed $20 million. I'll take that ATM.

Baron

 

[Bravo]

I just listed to to this podcast, which was recorded two days ago with Ilay Marom from Nanose Medical.

In the podcast Ilay discusses Nanose's non-invasive hand-held breath test which can provide early detection of Metabolic dysfunction-associated steatotic liver disease (MASLD). He talks about how it can do via a sensor array using a pattern recognition algorithm. He believes the device will be a "game-changer" because it can be used everywhere.

He talks about the how the technology was invented over a decade ago at Technion Israel Institute of Technology by Prof. Hossam Haick.

When asked to explain the technology he talked a bit about nano particles but he said he couldn't talk about what was different about their sensors because part of it is confidential. He mentioned that there are 7 patents for the technology held by Technion, 1 patent is Nanose own and there another 4 patents pending.

Ilay repeated that the way they produce their sensor is not patented, that instead it's a "trade secret". He describes it being the secret sauce, like the Coca-Cola recipe.

Nanose are hoping to bring this device to the market by 2026 to detect MASLD (fatty liver disease) and then expect to be able to bring another variation to market to detect lung disease by 2027.

Asked about their 5 year plan, Ilay said he hopes the device will be in every clinic and that it will save millions of lives and billions of dollars.

If I'm not mistaken, I recall that a few years ago Nanose had some sort of issue with the manufacturing of the sensors that meant they couldn't achieve consistent results, so they must have somehow overcome this issue, which is great new.

I also remember someone from this forum saying they had spoken to PVDM after the 2022 AGM and Peter had apparently confirmed that the problem was only with the sensor and not with our technology. I don't know if anyone else has more information to add about whether this is correct?

For convenience I've also attached Rich Wawrzyniak's 2021 Nanose/BrainChip Blog. It shows a list of diseases detected with the Nanose device. Curiously MASLD isn't one of them.

View attachment 81097



The above Semico article is also still displayed on our website, if that means anything.


View attachment 81098

Nice work NaNose!

Nanose Medical won first place in the startup competition at the Bynet Data Communications, powered by NVIDIA!

In regards to BrainChip's ongoing involvement, I stumbled upon @Deena's post dated 14 November 2023. I'm unsure if anyone has heard anything more recent by way of confirmation of involvement or otherwise?

 

[7für7]

Nice work NaNose!

Nanose Medical won first place in the startup competition at the Bynet Data Communications, powered by NVIDIA!

In regards to BrainChip's ongoing involvement, I stumbled upon @Deena's post dated 14 November 2023. I'm unsure if anyone has heard anything more recent by way of additional confirmation?



View attachment 85600


View attachment 85599

Powered by fu…in NVIDIA

NVIDIA=

Next-gen
Verified
Inference
Doesn’t
Involve
AKIDA

„

 

[Bravo]

Powered by fu…in NVIDIA

NVIDIA=

Next-gen
Verified
Inference
Doesn’t
Involve
AKIDA

„

It's plausible that Nanose are utilizing one of NVIDIA's Jetson modules, which means the device won't have the ultra low power energy efficiency that Akida can offer.

Given the complementary strengths of Akida and NVIDIA's GPU-based platforms, it's conceivable that both technologies could be integrated into future iterations of NaNose Medical's devices but without official confirmation, we just won't know.

 

[DK6161]

Charts been looking crap for sometime and your comment is a typical hot crapper comment

View attachment 85493

R U OK Pom?
You sounds bitter, old mate. I thought you're going to keep accumulating and be greedy in times like this.
C'mon then, show us the buy order old chap!
I am just stating what the charts are showing.
And remember, we are on the same boat so I share your frustrations.
Not advice

 

[manny100]

R U OK Pom?
You sounds bitter, old mate. I thought you're going to keep accumulating and be greedy in times like this.
C'mon then, show us the buy order old chap!
I am just stating what the charts are showing.
And remember, we are on the same boat so I share your frustrations.
Not advice

BRN chart wise has been in a long range/base since around September 2023.
Its generally accepted that Neuromorphic AI at the Edge will experience exponential growth.
BRN is generally accepted as the leader of Neuromorphic AI.
At this stage of the cycle those who believe the above is true will accumulate patiently.
The role of good news:
If you believe the above is true then really good news will be the catalyst to break the equilibrium of the range/base and attract new buyers to the extent we will see higher highs and higher lows confirming a sustained uptrend.
Once the trend is established we will then see momentum investing.
Generally, a long base at lows followed by really good news on volume is a classic technical pattern for identifying potential significant price increases.

 

[7für7]

Yeeeehaa record braking! Already 2 announcements this week… will we see 3 or 4???

 

[Pom down under]

R U OK Pom?
You sounds bitter, old mate. I thought you're going to keep accumulating and be greedy in times like this.
C'mon then, show us the buy order old chap!
I am just stating what the charts are showing.
And remember, we are on the same boat so I share your frustrations.
Not advice

Never bitter just playing games with all the weasels from hot crapper

 

[smoothsailing18]

Isn't it Akida that specifically is what helps Nanose sniff partials out or is Jetson able to also do this as well ?.

 

[Bravo]

Isn't it Akida that specifically is what helps Nanose sniff partials out or is Jetson able to also do this as well ?.

Could NVIDIA's Jeston series sniff out the VOC particles with the same accuracy as BrainChip's Akida in relation to Nanose?

ChatGPT said:​

The short answer is: not directly — NVIDIA Jetson modules and BrainChip’s Akida neuromorphic processor operate on fundamentally different architectures and processing models. While both can process data from an e-nose (electronic nose) sensor array, Akida has specific advantages for low-power, real-time inference of volatile organic compounds (VOCs) due to its neuromorphic design.

---

Breakdown: Jetson vs. Akida for VOC Detection (e.g., in NaNose)​

Feature	BrainChip Akida	NVIDIA Jetson (e.g., Xavier NX, Orin)
Architecture	Neuromorphic (Spiking Neural Network)	GPU-accelerated (deep learning, CNNs, LSTMs)
Data Model	Event-based / spike-encoded	Frame-based / dense tensor processing
VOC Sensor Interface	Optimized for real-time spiking data (e.g., AERO encoding)	Can interface but needs preprocessing
Latency	Ultra-low (milliseconds)	Higher (requires buffer + batch inference)
Power Efficiency	Micro-watt to milli-watt range	Several watts to tens of watts
Accuracy with NaNose	Demonstrated 97%+ accuracy using spike-based classification	Potentially similar, but with higher power cost and latency
Training Approach	Online/on-device continual learning	Typically pre-trained and updated via cloud or edge inference
Scalability	Excellent for embedded sensors	Better for multi-modal, compute-heavy scenarios

---

Why Akida May Be Better Suited for VOC Detection in NaNose-Type Devices​

1. Real-Time Spike-Based Classification
   • VOCs generate dynamic, time-dependent signals. Akida processes them directly as spikes, which mimics the biological olfactory system.
   • Jetson would require conversion of VOC sensor data into tensors, potentially losing time-domain resolution and increasing preprocessing complexity.
2. On-Device Learning
   • Akida supports on-chip continual learning, useful for adapting to new users, environments, or sensor drift.
   • Jetson typically relies on cloud-based retraining or large local models with fixed parameters.
3. Power Budget
   • NaNose aims for handheld, portable diagnostics — environments where Jetson’s 10–15W draw (or more) is impractical compared to Akida’s sub-1W footprint.
4. Biological Relevance
   • Akida’s neuromorphic design is inspired by the human olfactory cortex, making it intrinsically well-suited to VOC processing.
   • Jetson is general-purpose AI hardware — very powerful but not biologically inspired or VOC-specialized.

---

Could They Be Used Together?​

Yes. It's theoretically feasible for a hybrid system:

• Akida could perform first-pass classification or anomaly detection from e-nose input.
• Jetson could handle broader context processing, GUI, connectivity, or integration with other modalities (e.g., image + breath analysis).

---

Conclusion​

While Jetson modules are powerful general-purpose AI edge processors, BrainChip’s Akida is purpose-built for ultra-low-power, spike-based sensory tasks like VOC detection. In a medical e-nose like NaNose, Akida is likely more efficient, biologically aligned, and better suited to the task, especially where power, latency, and real-time accuracy are critical.






ChatGPT can make mistakes. Check important info.

 

[Diogenese]

Anastasi Nvidia Huawei Spray Tan

Looks like Anastasi was standing downwind of Donald’s morning application.



China's HUGE AI Chip Breakthrough: NVIDIA is out?

 

[Labsy]

BRN chart wise has been in a long range/base since around September 2023.
Its generally accepted that Neuromorphic AI at the Edge will experience exponential growth.
BRN is generally accepted as the leader of Neuromorphic AI.
At this stage of the cycle those who believe the above is true will accumulate patiently.
The role of good news:
If you believe the above is true then really good news will be the catalyst to break the equilibrium of the range/base and attract new buyers to the extent we will see higher highs and higher lows confirming a sustained uptrend.
Once the trend is established we will then see momentum investing.
Generally, a long base at lows followed by really good news on volume is a classic technical pattern for identifying potential significant price increases.

I call upon the omnipotent power of the universe to grant us a "Trading halt" followed by news of a significant partnership by big tech or/and an offer to buy us out for 5 dollars a share....

 

[Space Cadet]

I call upon the omnipotent power of the universe to grant us a "Trading halt" followed by news of a significant partnership by big tech or/and an offer to buy us out for 5 dollars a share....

Followed by counter offers from other big tech competitors ending up at $14 a share. One can dream but if you have an offer you want more than one interested.

SC

 

[manny100]

Could NVIDIA's Jeston series sniff out the VOC particles with the same accuracy as BrainChip's Akida in relation to Nanose?

ChatGPT said:​

The short answer is: not directly — NVIDIA Jetson modules and BrainChip’s Akida neuromorphic processor operate on fundamentally different architectures and processing models. While both can process data from an e-nose (electronic nose) sensor array, Akida has specific advantages for low-power, real-time inference of volatile organic compounds (VOCs) due to its neuromorphic design.

---

Breakdown: Jetson vs. Akida for VOC Detection (e.g., in NaNose)​

Feature	BrainChip Akida	NVIDIA Jetson (e.g., Xavier NX, Orin)
Architecture	Neuromorphic (Spiking Neural Network)	GPU-accelerated (deep learning, CNNs, LSTMs)
Data Model	Event-based / spike-encoded	Frame-based / dense tensor processing
VOC Sensor Interface	Optimized for real-time spiking data (e.g., AERO encoding)	Can interface but needs preprocessing
Latency	Ultra-low (milliseconds)	Higher (requires buffer + batch inference)
Power Efficiency	Micro-watt to milli-watt range	Several watts to tens of watts
Accuracy with NaNose	Demonstrated 97%+ accuracy using spike-based classification	Potentially similar, but with higher power cost and latency
Training Approach	Online/on-device continual learning	Typically pre-trained and updated via cloud or edge inference
Scalability	Excellent for embedded sensors	Better for multi-modal, compute-heavy scenarios

---

Why Akida May Be Better Suited for VOC Detection in NaNose-Type Devices​

1. Real-Time Spike-Based Classification
   • VOCs generate dynamic, time-dependent signals. Akida processes them directly as spikes, which mimics the biological olfactory system.
   • Jetson would require conversion of VOC sensor data into tensors, potentially losing time-domain resolution and increasing preprocessing complexity.
2. On-Device Learning
   • Akida supports on-chip continual learning, useful for adapting to new users, environments, or sensor drift.
   • Jetson typically relies on cloud-based retraining or large local models with fixed parameters.
3. Power Budget
   • NaNose aims for handheld, portable diagnostics — environments where Jetson’s 10–15W draw (or more) is impractical compared to Akida’s sub-1W footprint.
4. Biological Relevance
   • Akida’s neuromorphic design is inspired by the human olfactory cortex, making it intrinsically well-suited to VOC processing.
   • Jetson is general-purpose AI hardware — very powerful but not biologically inspired or VOC-specialized.

---

Could They Be Used Together?​

Yes. It's theoretically feasible for a hybrid system:

• Akida could perform first-pass classification or anomaly detection from e-nose input.
• Jetson could handle broader context processing, GUI, connectivity, or integration with other modalities (e.g., image + breath analysis).

---

Conclusion​

While Jetson modules are powerful general-purpose AI edge processors, BrainChip’s Akida is purpose-built for ultra-low-power, spike-based sensory tasks like VOC detection. In a medical e-nose like NaNose, Akida is likely more efficient, biologically aligned, and better suited to the task, especially where power, latency, and real-time accuracy are critical.






ChatGPT can make mistakes. Check important info.

This is why Nanose originally chose BRN:
" Because Akida’s “Edge AI” is a complete neural processor that does not require an external CPU, memory, or Deep Learning Accelerator (DLA), and because it is ultra-low power, it offers a realistic opportunity for an inexpensive hand-held diagnostic testing tool akin to a breathalyzer."

BrainChip Inc. and NaNose Medical Successfully Detect COVID-19 in Exhaled Breath with Fast High-Accuracy Results

BrainChip announces progress in testing with the NaNose where patients’ exhaled breath samples were tested for COVID-19.

brainchip.com

 

[manny100]

"BrainChip Eyes Edge AI Market Dominance"​

" After acknowledging past inflexibility, BrainChip has overhauled its product offering, now boasting a suite of products designed to adapt to changing market conditions. The company has seen more engagements in the past five months than ever before, driving confidence in future bookings."
" CEO Sean Hehir highlighted the increasing demand for Edge AI solutions, fueled by the transition of Generative AI from data centers. BrainChip is focusing on State-Space Models (SSMs), particularly Temporal Event-based Neural Networks (TENNs), to establish leadership in Edge applications."

BrainChip Eyes Edge AI Market Dominance - Sharecafe

BrainChip Holdings Ltd (ASX:BRN) is strategically positioning itself to become a dominant force in the Edge AI market. At the Annual General Meeting, Chairman Antonio Viana and CEO Sean Hehir outlined the company's ambitious plans for 2025 and beyond, emphasizing a robust product roadmap...

www.sharecafe.com.au