BRN Discussion Ongoing

[mcm]

Akida GenAI & Akida 3 have been adapted to handle 16-bit integer and 32-bit FP. This, in addition to the malleable architecture, enables these two chips to be flexibly configured to handle all types of models and to be adapted for future applications.

The provision of a LUT in place of an activation function seems like a patentable idea if original. We are also told by JT that a patent application is in the pipeline for a new technique for retrieving data from memory. This is the most energy intensive action so the invention will further increase the power efficiency and probably latency.

Akida 2 is 8 times more efficient than Akida 1, and presumably that also applies to GenAI & Akida 3 for equivalent Akida 1 tasks. However, 16-bit integer and 32-bit FP seem to provide excessive capabilities for an edge device. Does Nvidia need to look over its shoulder "like one that on a lonesome road doth walk in fear and dread, and having once turned round, walks on, and turns no more his head, because he knows that close behind a frightful fiend doth tread"?

Hey Diogenese,

Have you had a chance to look at what Nanovue is offering and whether or not it represents any serious competition to Akida? This is an interview the NVU CEO did with Stocks Down Under very recently in which he says "Having a very computational efficient processor that takes up very little space is very key in things like wearable glasses, medical devices, drones, putting it into cell phones ... anything that requires good battery management, also high computational power. EMASS right now benchmarks against the best of the best in those regards.":
It's the closest sounding tech to Akida I've come across however am not tech savvy enough to know how close it really is.
Cheers,
mcm

 

[CHIPS]

Hey Diogenese,

Have you had a chance to look at what Nanovue is offering and whether or not it represents any serious competition to Akida? This is an interview the NVU CEO did with Stocks Down Under very recently in which he says "Having a very computational efficient processor that takes up very little space is very key in things like wearable glasses, medical devices, drones, putting it into cell phones ... anything that requires good battery management, also high computational power. EMASS right now benchmarks against the best of the best in those regards.":
It's the closest sounding tech to Akida I've come across however am not tech savvy enough to know how close it really is.
Cheers,
mcm

They don't have any money either ... market cap 25,65 Millionen US-Dollar

Their market is normally totally different:

How the Company Makes Money
Nanoveu Ltd. generates revenue through the sale of its proprietary products and solutions. The company monetizes its technology by selling Nanoshield screen protectors directly to consumers as well as through distribution partnerships with retailers and device manufacturers. Additionally, the EyeFly3D product line is sold to consumers looking for enhanced visual experiences on their mobile devices. Revenue is further supported by licensing agreements and collaborations with technology partners looking to integrate Nanoveu's nanotechnology into their own products. Significant partnerships with manufacturers and retailers enhance the company's market reach and contribute to its earnings.

More: https://www.tipranks.com/stocks/au:nvu/stock-analysis

 

[Diogenese]

Hey Diogenese,

Have you had a chance to look at what Nanovue is offering and whether or not it represents any serious competition to Akida? This is an interview the NVU CEO did with Stocks Down Under very recently in which he says "Having a very computational efficient processor that takes up very little space is very key in things like wearable glasses, medical devices, drones, putting it into cell phones ... anything that requires good battery management, also high computational power. EMASS right now benchmarks against the best of the best in those regards.":
It's the closest sounding tech to Akida I've come across however am not tech savvy enough to know how close it really is.
Cheers,
mcm

It uses MRAM which implies analog.

Still in development.

From what I can see, nanoveu had no background in AI, but it recently bought EMASS.

https://company-announcements.afr.com/asx/nvu/b582a823-037d-11f0-9572-aafa4fbd818c.pdf

I think they bought a pig in a poke.

I'm not excited.

 

[mcm]

It uses MRAM which implies analog.

Still in development.

From what I can see, nanoveu had no background in AI, but it recently bought EMASS.

https://company-announcements.afr.com/asx/nvu/b582a823-037d-11f0-9572-aafa4fbd818c.pdf

I think they bought a pig in a poke.

I'm not excited.

Thanks Diogenese … I hope you’re right.
The only thing that makes me think it’s more than just a pig in a poke is that the CEO recently put up $350,000 of his own money to buy shares in a capital raise … and he has very impressive credentials. Now that is something I’ve never seen Sean or anyone else from BRN do.
Also Weebit Nano partnered with EMASS to demo hand gesture recognition at

 

[mcm]

Thanks Diogenese … I hope you’re right.
The only thing that makes me think it’s more than just a pig in a poke is that the CEO recently put up $350,000 of his own money to buy shares in a capital raise … and he has very impressive credentials. Now that is something I’ve never seen Sean or anyone else from BRN do.
Also Weebit Nano partnered with EMASS to demo hand gesture recognition at

Sorry, pressed the post button by mistake.
I was going to say they demonstrated with Weebit Nano at embedded world in Germany earlier this year showcasing hand gesture, recognition … which is something Brn has demonstrated on a number of occasions.
Anyway … hopefully Akita will hit the market well before them.

 

[Flenton]

Only just read the AGM address and my take on Sean's words but put into historic example is we are bring out this wonderful new technology. Its called a video cassette. Now, are we Beta or VHS?
Under Sean's leadership, Brainchip now has confirmation and validation from space and defence industries usage, akida being used in products including glasses for epilepsy seizure detection and in edge servers.
Brainchip is in a strong position to be the successful VHS.
There was a comment made years ago in an interview, I think, that Brainchip is trying to disrupt a market which didn't exist. Well, now it is appears to be small but it exists.

I've been seduced by their words before and I probably have been again but I do believe our day will come. That $9 million sales target or something like that will just be the beginning. Someone mentioned a few years ago that revenue will be lumpy and this could be the first lump.

I'll just keep holding and waiting. I've got time.

 

[Pom down under]

Hopefully we’re apart of there massive growth in sales in the next 3 years

 

[Rach2512]

Hopefully we’re apart of there massive growth in sales in the next 3 years

Great video thanks for sharing @Pom down under.

Here's a couple of screenshots.

Smart Health Monitoring Watch
Smart Video Door Bell
Indoor IOT Camera Alexa Speaker
And the list goes on.

Would be awesome if Akida was included in some of these products.

Also the last one reminds me of Ken.

 

[Rach2512]

Truly honored to have received the IEEE AESS Industrial Innovation Award… | Sandeep Gogineni | 31 comments

Truly honored to have received the IEEE AESS Industrial Innovation Award at this week’s IEEE International Radar Conference in Atlanta GA. Congratulations to the entire RFView team at Information Systems Laboratories whose contributions over the years have transformed RFView into a very...

www.linkedin.com

With an interesting congratulations.

 

[Jacob Kiraz]

Hopefully we’re apart of there massive growth in sales in the next 3 years

Weren't you next to Fact Finder at the AGM

 

[Deadpool]

I put this question in to the Chatster, and this answer popped out.

If BRN Akida is first to market with its low power ZSL tech, what do you predict the future will hold for BRN?



If BrainChip's Akida (BRN) is truly first to market with effective low-power zero-shot learning (ZSL) at the edge, it positions the company for major long-term strategic advantages across several industries. Here's a realistic breakdown of what the future could hold:

---

1. Early Mover Advantage​

• Implication: BRN could become the default platform for edge AI requiring fast adaptation, similar to how NVIDIA became dominant in GPU-accelerated AI.
• Outcome: Increased licensing deals, design wins in ASICs, and OEM partnerships, especially in automotive, defense, and finance.

---

2. Dominance in Edge AI Markets​

• Use Cases:
  • Wearables adapting to user behavior
  • Industrial IoT detecting anomalies in unseen environments
  • Autonomous vehicles identifying rare road scenarios
• Outcome: Akida becomes embedded in a wide range of smart devices needing intelligent, low-power, real-time decision-making.

---

3. Financial Upside​

• Short term: Stock could surge on early contracts and announcements.
• Medium term: Revenue growth from IP licensing, chips, and developer tools.
• Long term: Potential acquisition target for larger semiconductor or AI firms (e.g., Qualcomm, Intel, Apple).

---

4. Defensive Moat via Patents​

• Benefit: Early leadership allows BRN to file key patents around ZSL on neuromorphic hardware, creating barriers for competitors.
• Outcome: A strong IP portfolio could ensure royalties and licensing fees for years.

---

5. Risks Remain​

• Market adoption of neuromorphic chips is still early-stage
• Competitors like Intel (Loihi), SynSense, and IBM (TrueNorth) could catch up
• ZSL performance must be provably reliable in critical applications

---

Bottom Line:​

If BRN delivers and proves real-world ZSL performance at scale, it could become the ARM of neuromorphic AI—licensing essential IP across industries.

 

[Diogenese]

Thanks Diogenese … I hope you’re right.
The only thing that makes me think it’s more than just a pig in a poke is that the CEO recently put up $350,000 of his own money to buy shares in a capital raise … and he has very impressive credentials. Now that is something I’ve never seen Sean or anyone else from BRN do.
Also Weebit Nano partnered with EMASS to demo hand gesture recognition at

Please don't regard my ramblings as investment advice. I have no way to assess the commercial potential of competing applications, and I have an inherent bias in favour of BRN.

That said, I have looked at a few of the EMASS patents but don't really understand their "advantages" from a NN perspective.

It is worth noting the disclaimer re their power consumption in that it does not include the "peripheral" circuitry. This may exclude ADC and DAC circuits needed for the analog array to talk to the RISC-V CPU.

 

[GStocks123]

Maybe we’re still integrated with Edge Impulse/Qualcomm (docs updated 3 days ago)

Frequently asked questions (FAQ) | Edge Impulse Documentation

docs.edgeimpulse.com

 

[Diogenese]

Maybe we’re still integrated with Edge Impulse/Qualcomm (docs updated 3 days ago)

Frequently asked questions (FAQ) | Edge Impulse Documentation

docs.edgeimpulse.com

Hi GS,

While I think that the EI association will continue for a while, I don't see Qualcomm as a near term customer.

Qualcomm are committed to their inhouse DSP-based Hexagon 8 Ai solution. They distribute the AI workload between CPU/GPU/NPU depending on the type of work. Their NPU design philosophy is different from BRN's.

https://www.qualcomm.com/content/da...I-with-an-NPU-and-heterogeneous-computing.pdf

A personal assistant that offers a natural voice user interface (UI) to improve productivity and enhance user experiences is expected to be a popular generative AI application. The speech recognition, LLM, and speech models must all run with some concurrency, so it is desirable to split the models between the NPU, GPU, CPU, and the sensor processor. For PCs, agents are expected to run pervasively (always-on), so as much of it as possible should run on the NPU for performance and power efficiency.

...
Our latest Hexagon NPU offers significant improvements for generative AI, delivering 98% faster performance and 40% improved performance per watt. It includes micro-architecture upgrades, enhanced micro-tile inferencing, reduced memory bandwidth, and a dedicated power rail for optimal performance and efficiency. These enhancements, along with INT4 hardware acceleration, make the Hexagon NPU the leading processor for on-device AI inferencing.

...

5. Building our NPU from a DSP architecture was the right choice for improved programmability and the ability to tightly control scalar, vector, and tensor operations that are inherent to AI processing.

 

[7für7]

Maybe someone is able to read the technical details?

“news. May 9, 2025
"Dragonwing" equipped with Wi-Fi 7 compatible Japanese-made edge AI module
28th IoT Edge Computing EXPO
Silex Technology exhibited the EP-200Q, a system-on-module (SoM) for industrial edge AI equipped with Qualcomm's Dragonwing QCS6490 processor, at the 28th IoT/Edge Computing EXPO during Japan IT Week [Spring]. Aimed at battery-powered industrial/medical equipment.”

「Dragonwing」搭載　Wi-Fi 7対応の日本製エッジAIモジュール

サイレックス・テクノロジーは「Japan IT Week【春】」内「第28回 IoT・エッジコンピューティング EXPO」にて、Qualcommのプロセッサ「Dragonwing QCS6490」を搭載した産業用エッジAI向けのシステムオンモジュール（SoM）「EP-200Q」を展示した。バッテリー駆動の産業／医療機器に向ける。

eetimes.itmedia.co.jp

 

[Drewski]

The versions of this guy we have to contend with, both here and on the crapper are no doubt more degenerate in reality.

 

[GStocks123]

Hi GS,

While I think that the EI association will continue for a while, I don't see Qualcomm as a near term customer.

Qualcomm are committed to their inhouse DSP-based Hexagon 8 Ai solution. They distribute the AI workload between CPU/GPU/NPU depending on the type of work. Their NPU design philosophy is different from BRN's.

https://www.qualcomm.com/content/da...I-with-an-NPU-and-heterogeneous-computing.pdf

A personal assistant that offers a natural voice user interface (UI) to improve productivity and enhance user experiences is expected to be a popular generative AI application. The speech recognition, LLM, and speech models must all run with some concurrency, so it is desirable to split the models between the NPU, GPU, CPU, and the sensor processor. For PCs, agents are expected to run pervasively (always-on), so as much of it as possible should run on the NPU for performance and power efficiency.

...
Our latest Hexagon NPU offers significant improvements for generative AI, delivering 98% faster performance and 40% improved performance per watt. It includes micro-architecture upgrades, enhanced micro-tile inferencing, reduced memory bandwidth, and a dedicated power rail for optimal performance and efficiency. These enhancements, along with INT4 hardware acceleration, make the Hexagon NPU the leading processor for on-device AI inferencing.

...

5. Building our NPU from a DSP architecture was the right choice for improved programmability and the ability to tightly control scalar, vector, and tensor operations that are inherent to AI processing.

Appreciate your input Dio

 

[Fullmoonfever]

Maybe we’re still integrated with Edge Impulse/Qualcomm (docs updated 3 days ago)

Frequently asked questions (FAQ) | Edge Impulse Documentation

docs.edgeimpulse.com

Appears they may be talking about still supporting "existing trained Edge Impulse projects" but still can't train new models.

Does stiffle anyone wanting to try train a new model for Akida which is BS imo but is Qualcomms want I guess.

I think someone posted previously that Tony Lewis (?) said it wasn't an issue and something about contract review or similar. Wouldn't expect it should take this long for a review though if all good?

BrainChip AKD1000 | Edge Impulse Documentation

docs.edgeimpulse.com

Update March 25, 2025: At this time the training of Brainchip models is suspended. You may still use existing trained Edge Impulse projects to deploy to Brainchip devices. Please make a post on https://forum.edgeimpulse.com/ or contact your support representative if you need assistance at this time.

 

[Diogenese]

Appears they may be talking about still supporting "existing trained Edge Impulse projects" but still can't train new models.

Does stiffle anyone wanting to try train a new model for Akida which is BS imo but is Qualcomms want I guess.

I think someone posted previously that Tony Lewis (?) said it wasn't an issue and something about contract review or similar. Wouldn't expect it should take this long for a review though if all good?

BrainChip AKD1000 | Edge Impulse Documentation

docs.edgeimpulse.com

Update March 25, 2025: At this time the training of Brainchip models is suspended. You may still use existing trained Edge Impulse projects to deploy to Brainchip devices. Please make a post on https://forum.edgeimpulse.com/ or contact your support representative if you need assistance at this time.

In contract negotiations the number of opinions is proportional to the square of the number of lawyers.

 

[CHIPS]

I put this question in to the Chatster, and this answer popped out.

If BRN Akida is first to market with its low power ZSL tech, what do you predict the future will hold for BRN?



If BrainChip's Akida (BRN) is truly first to market with effective low-power zero-shot learning (ZSL) at the edge, it positions the company for major long-term strategic advantages across several industries. Here's a realistic breakdown of what the future could hold:

---

1. Early Mover Advantage​

• Implication: BRN could become the default platform for edge AI requiring fast adaptation, similar to how NVIDIA became dominant in GPU-accelerated AI.
• Outcome: Increased licensing deals, design wins in ASICs, and OEM partnerships, especially in automotive, defense, and finance.

---

2. Dominance in Edge AI Markets​

• Use Cases:
  • Wearables adapting to user behavior
  • Industrial IoT detecting anomalies in unseen environments
  • Autonomous vehicles identifying rare road scenarios
• Outcome: Akida becomes embedded in a wide range of smart devices needing intelligent, low-power, real-time decision-making.

---

3. Financial Upside​

• Short term: Stock could surge on early contracts and announcements.
• Medium term: Revenue growth from IP licensing, chips, and developer tools.
• Long term: Potential acquisition target for larger semiconductor or AI firms (e.g., Qualcomm, Intel, Apple).

---

4. Defensive Moat via Patents​

• Benefit: Early leadership allows BRN to file key patents around ZSL on neuromorphic hardware, creating barriers for competitors.
• Outcome: A strong IP portfolio could ensure royalties and licensing fees for years.

---

5. Risks Remain​

• Market adoption of neuromorphic chips is still early-stage
• Competitors like Intel (Loihi), SynSense, and IBM (TrueNorth) could catch up
• ZSL performance must be provably reliable in critical applications

---

Bottom Line:​

If BRN delivers and proves real-world ZSL performance at scale, it could become the ARM of neuromorphic AI—licensing essential IP across industries.

I asked GROK the same question and the bottom line here is:

Conclusion
BrainChip’s Akida, with its low-power, ZSL-like neuromorphic capabilities, positions it as a pioneer in edge AI. If it sustains its first-mover advantage, it could dominate niche markets like space and defense, achieve moderate success in automotive and IoT, or become an acquisition target. However, competition, adoption risks, and economic factors temper optimism. By 2030, BrainChip is likely to be a notable player in edge AI, with success hinging on execution and market demand for its unique technology.