BRN Discussion Ongoing

[7für7]

#gsa #whereleadermeet #semiconductor #technology #ai #globalcollaboration | Global Semiconductor Alliance

🚨ONE WEEK.🚨 The countdown is on: The GSA U.S. Executive Forum returns to the Rosewood Sand Hill next week. 200+ global executives. 20 powerhouse speakers. One unforgettable program. Spots are almost gone—secure your place today before registration closes: https://lnkd.in/gbK469Ai See you on...

lnkd.in

View attachment 91313
View attachment 91314

Happy as Larry

BrainChip as the Networking Break Sponsor? Well, everyone knows what usually happens to opening bands…

Wohoooo GO BRAINCHIP!!!!

 

[Xray1]

Looks like someone has achieved their goal of placing Akida in a well-known online provider… if that doesn’t scream for an unquoted securities announcement, I don’t know what does!

DYOR

I wonder if this will count as one of our CEO's so called ... anticipated "Booking" !!!??

 

[Fullmoonfever]

Neuromorphic article recently posted on Japanese Media site Note.

Some snips on Akida.

ニューロモーフィックコンピューティング：脳の神経回路を模倣した次世代AI技術の深層｜sabosaboten32

はじめに：AI技術の隠れた革命 生成AIブームの影で、静かに進歩を続ける革新的な技術があります。それが「ニューロモーフィックコンピューティング」です。 現在のAI技術は、デジタル信号で動作するフォン・ノイマン型コンピュータ上で実行されています。しかし、人間の脳は全く異なる方式で動作します。脳神経細胞（ニューロン）は、スパイク状の電気信号を断続的に発生させ、極めて低い消費電力で高度な情報処理を実現しています。 AI技術コンサルタントとして、エッジAI実装プロジェクトを数多く手がけてきた経験から、この技術が持つ革命的な可能性と、実用化への道筋について詳しく解説します。 1. ニュー

note.com

Neuromorphic computing: A deep layer of next-generation AI technology that mimics the brain's neural circuits

Sabosaboten32
September 16, 2025 18:45

Power consumption during inference processing (per 1000 inferences):
NVIDIA A100 GPU: ~280W
Intel CPU (Xeon): ~150W
ARM Cortex-M4: ~50mW
Intel Loihi 2: ~1mW
IBM TrueNorth: ~70mW
Brainchip Akida: ~200µW

BrainChip Akida: Commercial Neuromorphic SoC​

Technological innovation​

• One-Shot Learning function
• Compatible with Edge Learning
• 28nm FD-SOI manufacturing
• Power consumption: 200µW～2mW

Practical application example: Automotive field​

ADAS(Advanced driver assistance system) implementation:
- Object detection accuracy: 92.4%
- Latency: < 1ms
- Power consumption: 600µW
- Operating temperature range: -40℃ ～ +125℃

# Building a neuromorphic development environment
1. Simulation Phase
- Model verification with Brian2, NEST, and BindsNET

2. Hardware Mapping
- Loihi: Nx SDK
- SpiNNaker: sPyNNaker
- Akida: MetaTF

3. Performance Tuning
- Spike rate optimization
- Network partitioning strategy
- Memory usage adjustment


References/technical materials:

• Intel Loihi Technical Documentation
• IBM TrueNorth Research Papers
• BrainChip Akida Development Guide
• Nature Neuromorphic Computing Review
• IEEE Neuromorphic Engineering Conference Proceedings

 

[GStocks123]

Worth the read! Akida used in a few different studies.

https://scholarsmine.mst.edu/cgi/viewcontent.cgi?article=4429&context=doctoral_dissertations

 

[Esq.111]

Afternoon Chippers ,

May see some large orders , or not , After Market today ........

Quadruple Witching Day ,

Rebalancing of portfolios , 3rd Friday in September , Today.

Regards ,
Esq.

 

[7für7]

Afternoon Chippers ,

May see some large orders , or not , After Market today ........

Quadruple Witching Day ,

Rebalancing of portfolios , 3rd Friday in September , Today.

Regards ,
Esq.

And don’t forget… it’s the last 19th September of this year…

 

[Space Cadet]

Help Dodgy Knees,

https://www.eetimes.com/sifive-unleashes-second-generation-risc-v-ip/

 

[TECH]

Nice work Steve B.

Marketing, promoting, and most of all, opening up doors to the wider community, our footprint is expanding, as it should be!

DigiKey is another huge step into the commercial unknown, but without these branches we will never grow.

Slow but positive steps forward, keep feeding and watering our Akida dream, on behalf of our loyal Australian shareholder base, I say thank you to our entire team.

Akida Tech.

 

[Tothemoon24]

 

[Pom down under]

 

[Pom down under]

 

[CHIPS]

S

Careful, it is old! It can‘t be from 2025 because some information is outdated.

 

[Pom down under]

S

Careful, it is old! It can‘t be from 2025 because some information is outdated.

Never even listened too it, but it does say 2025 August so can’t be old?

 

[yogi]

Akida Cloud from BrainChip lets companies stream real data from their devices directly to physical FPGAs running Akida IP - all through a browser.

Why does this matter? Companies can now test their AI models with real-world data on the actual hardware that would go into production ASICs, without the hassle and cost of procuring their own FPGA development boards.

#ai #edgeai #machinelearning #fpga #aihardware #arvr | Kurt Manninen

Akida Cloud from BrainChip lets companies stream real data from their devices directly to physical FPGAs running Akida IP - all through a browser. Why does this matter? Companies can now test their AI models with real-world data on the actual hardware that would go into production ASICs...

www.linkedin.com

 

[smoothsailing18]

S

Careful, it is old! It can‘t be from 2025 because some information is outdated.

They should get up to speed with the reality we are already commercialised.

 

[MDhere]

Never even listened too it, but it does say 2025 August so can’t be old?

Yep thanks Pom, listened to it, they talked about the concept of Brn, low power, on edge, llms etc etc. The guy on there, his voice almost sent me back to sleep.
One thing I didn't like which woke me back up was he says its a big IF. it is NOT a big IF so it's clear the guy has put together a vague script or even just improvised to fill the minutes up.

Overall, a fair extremely vague overview not much depth, more like flipping a pancake back and forth on a pan to get the right outcome in the guys delivery of words.

At least a plug for Brainchip on MF channel, (which surprises me) , Thanks for the post Pom. I have listened to it for you. lol

 

[Pom down under]

Yep thanks Pom, listened to it, they talked about the concept of Brn, low power, on edge, llms etc etc. The guy on there, his voice almost sent me back to sleep.
One thing I didn't like which woke me back up was he says its a big IF. it is NOT a big IF so it's clear the guy has put together a vague script or even just improvised to fill the minutes up.

Overall, a fair extremely vague overview not much depth, more like flipping a pancake back and forth on a pan to get the right outcome in the guys delivery of words.

At least a plug for Brainchip on MF channel, (which surprises me) , Thanks for the post Pom. I have listened to it for you. lol

Thanks for listening to it for me and the update. Have a lovely weekend

Pom

 

[Frangipani]

Big thanks to Kurt Manninen for showcasing the impressive Akida Cloud RGB gesture recognition demo 😄!! | BrainChip

Big thanks to Kurt Manninen for showcasing the impressive Akida Cloud RGB gesture recognition demo 😄!!

www.linkedin.com

 

[Frangipani]

#snn #neuromorphic | Low Kelvin

Unboxing the future of embedded AI with the Akida AKD1000 SNN chip – true neuromorphic hardware that supports event-based SNN architecture with 1.2m neurons over 80 neural cores, capable of on-chip incremental learning (for the last layer) on the edge. Very glad to have received this as a gift...

www.linkedin.com

Temasek (company) - Wikipedia

en.m.wikipedia.org

 

[itsol4605]