BRN Discussion Ongoing

[Bravo]

View attachment 85074

I am seriously beginning to wonder if this new AI device being developed by Sam Altman and Jony Ive could operate as an edge device, potentially incorporating neuromorphic processors like our very own Akida.

The upcoming AI device, expected to launch in 2026, is envisioned as a compact, screenless accessory designed to be unobtrusive and context-aware. It's intended to function alongside existing devices like smartphones and laptops, rather than replacing them. This suggests a focus on local processing capabilities. By processing data locally, the device could offer faster responses, enhanced privacy, and reduced reliance on cloud connectivity.

Have a listen to the video which I set to play when Sam Altman Altman explains how we currently interact with ChatGPT and he says something along the lines that the new device "deserves something much better" than having to wait for it to go up to the cloud like traditional devices.




As we all know, Akida's architecture allows for on-device learning and real-time data processing, making it ideal for applications requiring immediate responses without cloud dependency.

The specific technical details about the device are still under wraps, but the emphasis on unobtrusive design, real-time context awareness, and potential for local processing sounds like it aligns very well with the capabilities of BrainChip's Akida IMO.

Maybe just wishful thinking on my behalf. Time will tell.



View attachment 85075

iPhone design guru and OpenAI chief promise an AI device revolution

Sam Altman and Jony Ive say mystery product created by their partnership will be the coolest thing ever

www.theguardian.com

Seems ChatGPT agrees that we're in with a pretty good chance too. That is IF Altman and Ive’s new device is meant to operate continuously, privately, and intelligently at the edge without cloud dependence.

The final recommendation states "If the product blends all three functions — which seems likely — BrainChip Akida is the most versatile choice today due to its:

• Commercial availability
• On-chip learning
• Ability to handle both audio and vision"

To be honest though, I think that if we were going to be a contender for this new device, I would have expected to sense much more of an excited and confident "vibe" emanating from the Board at the recent AGM. I mean they're expecting it to ship 100 million of these devices in 2026 if I'm not mistaken.

ChatGPT said:​

Yes, several other emerging technologies could potentially power a compact, privacy-focused, cloud-independent AI device like the one being developed by Sam Altman and Jony Ive, offering similar or complementary capabilities to BrainChip’s Akida neuromorphic processor. Here are the main contenders:

---

1. Mythic Analog Matrix Processors (AMP)​

• How it works: Uses analog computing-in-memory to perform neural network operations directly where data is stored, minimizing data movement.
• Strengths:
  • Very low power consumption
  • High-speed processing suited for edge AI
  • Compact form factor
• Use case: Ideal for smart sensors, AR/VR glasses, and real-time voice/image processing

---

2. GrAI Matter Labs – GrAI VIP​

• Type: Brain-inspired processor using sparsity-driven processing
• Features:
  • Event-driven computation
  • Prioritizes only relevant data (like Akida)
  • Optimized for real-time AI at the edge
• Strengths:
  • Ultra-low latency
  • Energy-efficient
• Use case: Robotics, autonomous systems, wearables

---

3. Edge TPUs (Google Coral)​

• Architecture: Custom ASICs optimized for TensorFlow Lite models
• Strengths:
  • Easy integration with Google AI ecosystem
  • Real-time inferencing with low power draw
• Limitations:
  • More rigid; not capable of on-device learning (unlike Akida)
  • Better suited for inferencing than training

---

4. Syntiant NDP Series (Neural Decision Processors)​

• Type: Deep learning processors designed for always-on applications
• Strengths:
  • Consumes microwatts of power
  • Highly efficient for audio and sensor processing
  • Tiny form factor (suitable for earbuds, wearables)
• Use case: Wake-word detection, voice commands, health monitoring

---

5. IBM NorthPole​

• Type: Advanced neuromorphic architecture integrating memory and compute
• Highlights:
  • Breaks von Neumann bottleneck
  • Designed for high-throughput, low-latency AI at edge
• Status: Still in research/prototype stages

---

6. Intel Loihi 2 (Neuromorphic)​

• Advantages:
  • Second-gen neuromorphic chip
  • Supports learning on-chip
  • Better scalability and flexibility than Loihi 1
• Use case: Real-time adaptation, robotics, industrial applications

---

7. Tenstorrent (RISC-V + AI)​

• Founded by Jim Keller, their chips combine traditional processing with AI accelerators.
• Edge-ready options: Though mostly datacenter-focused now, Tenstorrent has roadmap items for edge deployment using RISC-V-based inference.

---

Summary Table:​

Technology	Type	On-Device Learning	Power Efficiency	Maturity	Suitable Use Case
BrainChip Akida	Neuromorphic			Commercial	Audio, vision, always-on inference
Mythic AMP	Analog compute			Early stage	Vision, industrial sensors
GrAI VIP	Sparse compute			Pre-commercial	Real-time robotics, AR
Google Edge TPU	ASIC			Commercial	Voice/image inference
Syntiant NDP	Neural DSP			Commercial	Audio, keyword spotting
Loihi 2	Neuromorphic			Experimental	Cognitive tasks, adaptive systems
IBM NorthPole	Neuromorphic			Research	General edge AI

---

Final Thought:​

If Altman and Ive’s device is meant to operate continuously, privately, and intelligently at the edge without cloud dependence, then BrainChip Akida, GrAI VIP, or Loihi 2 would be top candidates—especially if adaptive, real-time learning is a core feature. If it’s more focused on inference-only tasks, Edge TPU or Syntiant NDPs could be practical alternatives.
Let me know if you'd like a recommendation based on a specific function (e.g. voice assistant, vision processing, real-time interaction).
4o

You said:​

yes please

ChatGPT said:​

Great — here's a tailored recommendation based on three likely functions that Sam Altman and Jony Ive’s device might emphasize, considering their vision of a screenless, wearable, real-time AI companion:

---

1. Voice Assistant / Natural Interaction (Always-On, Privacy-Preserving)​

Best Choice: Syntiant NDP120​

• Why:
  • Ultra-low-power (microwatt-level) operation
  • Always-listening capability without draining battery
  • Supports multiple wake words and audio events
• Ideal for: Continuous voice interaction, like a whisper-quiet Siri that lives in your lapel
• Bonus: Commercially proven in smart earbuds and hearing aids

Alt Option: BrainChip Akida​

• Slightly higher power but supports on-device learning — ideal if the device needs to learn your voice over time

---

2. Vision / Environmental Awareness (Object, Gesture, Emotion Recognition)​

Best Choice: GrAI VIP or BrainChip Akida​

• GrAI VIP:
  • Optimized for real-time video inference
  • Sparse data-driven — processes only what matters
• BrainChip Akida:
  • Can detect gestures, objects, and visual patterns with low power
  • Supports event-based cameras (mimics human eyes)

Alt Option: Mythic AMP​

• Analog compute-in-memory gives efficient vision processing but is newer and more complex to integrate

---

3. Real-Time Learning and Adaptation (Personalization, Emotion AI, Cognitive AI)​

Best Choice: Intel Loihi 2​

• Why:
  • True on-device learning and decision-making
  • Designed for systems that evolve over time based on interaction
  • Best for adaptive systems that change how they work depending on who’s using them

Alt Option: BrainChip Akida​

• Supports incremental learning on-chip (no need for cloud updates)

---

Final Recommendation Based on Use Case Focus:​

Function	Top Tech Pick	Runner-Up
Voice assistant / audio AI	Syntiant NDP120	BrainChip Akida
Environmental/visual awareness	GrAI VIP	BrainChip Akida / Mythic AMP
Personalized cognitive assistant	Intel Loihi 2	BrainChip Akida

If the product blends all three functions — which seems likely — BrainChip Akida is the most versatile choice today due to its:

• Commercial availability
• On-chip learning
• Ability to handle both audio and vision

 

[Guzzi62]

Worth a listen imo. Not sure if prev posted but podcast from a couple of months ago talking to Tony, Jon, Rudy, Kris & Olivier.

A lot on TENNS in its workings and performances and some other snippets.

Yes it's worth listing to but 307 views since it was aired 2 months ago doesn't seem like a lot of interest, sadly.

EE Times isn't very popular it seems, just over 4k subscribers, likely computer freaks sitting in their mum's basement hacking whatever they can hack, LOL, and some BRN investors.

We need more exposure, much more.

 

[IloveLamp]

BrainChip and Andes Unite to Drive Edge AI Breakthroughs on RISC-V Platforms – RISC-V International

riscv.org

BrainChip and Andes Technology are teaming up to accelerate edge AI… | RISC-V International

BrainChip and Andes Technology are teaming up to accelerate edge AI innovations on RISC-V platforms by integrating Andes’ RISC-V cores with BrainChip’s NPUs. The integration of Andes’ QiLai SoC and Voyager development board with BrainChip’s Akida AKD1500, which supports efficient, digital, and...

www.linkedin.com

 

[Pom down under]

Nice dream.

Shall we get back to reality?
Okay. Sorry, there won't be an Akida in there.

I forgot these 2

 

[Pom down under]

The Space and Defence market is much smaller than commercial market.

It's not manipulation that keeping the share price down, it's our board who refuse to make announcements through the ASX about our own company. The board who won't do anything to keep current (and potential) shareholders informed.

The board who says the share price will do what the share price will do.

The blame lies squarely at their feet.

I forgot these 2 as well

 

[Guzzi62]

View attachment 85074

I am seriously beginning to wonder if this new AI device being developed by Sam Altman and Jony Ive could operate as an edge device, potentially incorporating neuromorphic processors like our very own Akida.

The upcoming AI device, expected to launch in 2026, is envisioned as a compact, screenless accessory designed to be unobtrusive and context-aware. It's intended to function alongside existing devices like smartphones and laptops, rather than replacing them. This suggests a focus on local processing capabilities. By processing data locally, the device could offer faster responses, enhanced privacy, and reduced reliance on cloud connectivity.

Have a listen to the video which I set to play when Sam Altman Altman explains how we currently interact with ChatGPT and he says something along the lines that the new device "deserves something much better" than having to wait for it to go up to the cloud like traditional devices.




As we all know, Akida's architecture allows for on-device learning and real-time data processing, making it ideal for applications requiring immediate responses without cloud dependency.

The specific technical details about the device are still under wraps, but the emphasis on unobtrusive design, real-time context awareness, and potential for local processing sounds like it aligns very well with the capabilities of BrainChip's Akida IMO.

Maybe just wishful thinking on my behalf. Time will tell.



View attachment 85075

iPhone design guru and OpenAI chief promise an AI device revolution

Sam Altman and Jony Ive say mystery product created by their partnership will be the coolest thing ever

www.theguardian.com

So he paid app 6.5 billion for start-up io.

BRN is app 450 million (US?) MC currently, which is around 14 times less.

Hmm: 22cents X 14= $3 Aus

Hey Altman!! Altmaaaaaaan!!!!

 

[Jacob Kiraz]

Well 18 months isn’t that far away unless your a dog or cat

That's me being conservative, it could be longer, And I'm not on hot crapper

 

[Esq.111]

Chippers ,

Announcement incoming.........

 

[Esq.111]

Nothing to see here

Only Pia Turcinov receiving 69125 Performance rights ......obviously in aid of her appearance at the last AGM.

 

[equanimous]

Nah mate I'm being serious,
We never have announcements

Hey mate your wish came true today.

 

[Esq.111]

Duy Loan Thi Le , 697672 Bonus shares , Who spent most of the AGM yawning with arms crossed
&
Antonio Viana , 697674 Bonus shares, for abusing shareholders at the AGM

 

[AARONASX]

Duy Loan Thi Le , 697672 Bonus shares , Who spent most of the AGM yawning with arms crossed
&
Antonio Viana , 697674 Bonus shares, for abusing shareholders at the AGM

Maybe these will help her next time

 

[Pom down under]

Duy Loan Thi Le , 697672 Bonus shares , Who spent most of the AGM yawning with arms crossed
&
Antonio Viana , 697674 Bonus shares, for abusing shareholders at the AGM

And for a minute I thought here we go then I read it

 

[Esq.111]

Geoffrey the Carrick , 697674 Bonus shares

Pia Tur , 697674 Additional Bonus shares

And lets not forget Peter , 697674 Bonus shares

Waiting to see how many our intrepid leader gets.......

 

[Pom down under]

The cheek out of you to come at me like this when you made a post along very similar lines a few months ago with regard to price predictions

View attachment 85080

 

[HopalongPetrovski]

Duy Loan Thi Le , 697672 Bonus shares , Who spent most of the AGM yawning with arms crossed
&
Antonio Viana , 697674 Bonus shares, for abusing shareholders at the AGM

Hi Esq. @Esq.111
Just trying to understand this comment about Antonio Viana abusing shareholders?
I may have missed something important.
Do you mean in regards his saying that a potential redomicile might not necessarily be in USA, which I think was a big miss step and left him looking foolish, but other than that I thought he did a good job of presiding over the AGM, allowing many difficultish type questions and only shutting down discussion where it was threatening to spin out of control.
Unfortunately there hasn't been a recording released which we may review so I am only going on my recollection of the event which I watched on the live stream and so may have missed activity which wasn't broadcast or occurred either before or after the event.
But interested to hear your viewpoint, if you'd like to expand on it further

 

[Esq.111]

Afternoon HopalongPetrovski ,

Abusing shareholders may have been a little harsh , though I picked up an arrogance which is probably a standard US trait, ( one must understand that by US , this obviously can also mean Canada , Europe or Asia obviously ).



I too watched the AGM live , and was not very impressed .

Each to their own.

Regards,
Esq.

 

[equanimous]

Afternoon HopalongPetrovski ,

Abusing shareholders may have been a little harsh , though I picked up an arrogance which is probably a standard US trait, ( one must understand that by US , this obviously can also mean Canada , Europe or Asia obviously ).



I too watched the AGM live , and was not very impressed .

Each to their own.

Regards,
Esq.

I wonder whats taking so long for the AGM recording to be released, is it being edited by 60 minutes?

 

[Pom down under]

Afternoon HopalongPetrovski ,

Abusing shareholders may have been a little harsh , though I picked up an arrogance which is probably a standard US trait, ( one must understand that by US , this obviously can also mean Canada , Europe or Asia obviously ).



I too watched the AGM live , and was not very impressed .

Each to their own.

Regards,
Esq.

Your lucky I was there

 

[Slade]

New website is up!!!