BRN Discussion Ongoing

[Dallas]

Neuromorphic Computing: Future of Brain-Inspired AI Technology

Explore how Neuromorphic Computing is redefining AI with brain-inspired chips, ultra-low power consumption, and real-time learning for edge AI applications.

exelanceit.in

 

[Guzzi62]

From Dallas's link above!

Neuromorphic Computing: Revolutionizing the Future of AI with Brain-Inspired Technology​

Exelance IT > Blogs > IT Industry > Neuromorphic Computing: Revolutionizing the Future of AI with Brain-Inspired Technology

• ExeLance IT - Author
• December 19, 2024
• No Comments

Picture a future where artificial intelligence seamlessly integrates into our daily lives, mimicking the efficiency and adaptability of the human brain. This isn’t science fiction—it’s the promise of neuromorphic computing, a revolutionary approach that’s set to transform the landscape of AI as we know it.
But why is this technology causing such a stir in the tech world? Traditional computing systems, despite their impressive capabilities, pale in comparison to the human brain’s energy efficiency and processing power. Enter neuromorphic computing—a game-changing solution that draws inspiration from the very organ it aims to emulate. By designing chips that function more like biological neural networks, researchers are unlocking new possibilities for AI that are faster, more efficient, and incredibly powerful.

As we delve into the world of neuromorphic computing, we’ll explore its fundamental principles, the groundbreaking chips making waves in edge AI, and spotlight innovative solutions like BrainChip’s Akida and TENN. Join us on this journey to uncover how brain-inspired technology is revolutionizing the future of AI and paving the way for smarter, more intuitive machines.

Neuromorphic Approach to AI​

Understanding Neuromorphic Computing​

Neuromorphic computing is an innovative approach to artificial intelligence that mimics the brain’s neural networks. This field aims to create systems that potentially revolutionize AI and machine learning.

Key Principles of Neuromorphic Computing​

• Parallel Processing: Like the brain, neuromorphic systems process information in parallel, enabling faster and more efficient computations.
• Low Power Consumption: Designed to be energy-efficient, similar to the human brain’s remarkable efficiency.
• Adaptive Learning: Can learn and adapt in real-time, like biological neural networks.

Feature	Traditional Computing	Neuromorphic Computing
Architecture	Sequential	Parallel
Power Efficiency	High consumption	Low consumption
Learning Ability	Predefined algorithms	Adaptive, real-time learning
Data Processing	Separate memory and processing	Integrated memory and processing

Traditional vs Neuromorphic Computing

Advantages of Neuromorphic AI​

• Enhanced pattern recognition
• Improved natural language processing
• Real-time decision-making capabilities
• Reduced latency in edge computing
• More human-like AI interactions

Neuromorphic Chips and Edge AI​

Revolutionizing Edge AI with Neuromorphic Chips​

Neuromorphic chips are transforming edge AI, particularly with their advantages in power consumption and real-time processing.

Key Benefits of Neuromorphic Chips​

• Energy Efficiency
• Low Latency
• On-device Learning
• Adaptability

Feature	Traditional Chips	Neuromorphic Chips
Power Consumption	High	Ultra-low
Processing Speed	Variable	Near real-time
Learning Capability	Limited	Continuous
Adaptability	Fixed	Dynamic

Traditional Chips vs Neuromorphic Chips

Applications in Edge Devices​

• Smart sensors for industrial IoT
• Autonomous vehicles
• Wearable health monitors
• Intelligent home assistants

BrainChip’s Akida and TENN​

Akida: A Neuromorphic System-on-Chip​

BrainChip’s Akida mimics the brain’s neural structure, offering ultra-low power consumption and real-time edge AI capabilities.

Key Features of Akida​

• Neural Network Processor for efficient processing
• On-chip learning without cloud connectivity
• Event-based processing for low latency
• Flexible architecture supporting CNNs and SNNs

TENN: Temporal Event Neural Network​

TENN is BrainChip’s proprietary architecture designed for processing temporal data efficiently with improved accuracy.

Applications of Akida and TENN​

• Autonomous vehicles
• Industrial IoT
• Smart home devices
• Healthcare monitoring systems
• Cybersecurity

Conclusion​

Neuromorphic computing is revolutionizing AI by mimicking the brain’s efficiency and adaptability. Technologies like BrainChip’s Akida and TENN are driving smarter, energy-efficient edge AI solutions that pave the way for faster, more intuitive, and intelligent systems in our daily lives.

 

[Esq.111]

Good Afternoon Chippers ,

Very quiet on the volume front thus far , hour of power into the close ..... then late trades should be interesting.

CRANK IT. All good ... Time stamp 34.29




Also , on a side note , ASX trading hours days closed .

Trading calendars​

2024​

2025​

>

2024 Trading calendar​

PUBLIC HOLIDAY	DATES FOR 2024	TRADING DAY[1]	SETTLEMENT ACTIVITY – ASX Settlement (CHESS)	SETTLEMENT ACTIVITY – ASX Clear (Derivatives)	BUSINESS DAY[2]
New Year’s Day	Monday 1 January	CLOSED	No Settlement	No Settlement	NO
Australia Day	Friday 26 January	CLOSED	No Settlement	No Settlement	NO
Good Friday	Friday 29 March	CLOSED	No Settlement	No Settlement	NO
Easter Monday	Monday 1 April	CLOSED	No Settlement	No Settlement	NO
ANZAC Day	Thursday 25 April	CLOSED	No Settlement	No Settlement	NO
King’s Birthday	Monday 10 June	CLOSED	No Settlement	No Settlement	NO
Last Business day before Christmas Day	Tuesday 24 December	CLOSE EARLY[3]	Settlement	Settlement	YES
Christmas Day	Wednesday 25 December	CLOSED	No Settlement	No Settlement	NO
Boxing Day	Thursday 26 December	CLOSED	No Settlement	No Settlement	NO
Last Business day of the Year	Tuesday 31 December	CLOSE EARLY[4]	Settlement	Settlement	YES

Best Wishes All .
Esq.

 

[Esq.111]

Something has started to play with the BUY side

1,635,278 units @ $0.25

 

[Diogenese]

PROPHESEE | Metavision Technologies

REVEAL THE INVISIBLE with the world's most advanced neuromorphic vision system, inspired by human vision and built on the foundation of neuromorphic engineering. PROPHESEE gives Metavision to machines, revealing what was previously invisible to them.

www.prophesee.ai

What are prophesee currently using that allows 2mW sensing power consumption?

Also good to see they have brainchip listed as a partner. Synsense who they have previously partnered with arent listed

Synsense have merged with Inivation, a rival DVS maker.

https://www.synsense.ai/synsense-an...orm-leading-neuromorphic-technology-provider/


Not sure where you got the 2 mW figure, but does it relate to their newish GenX320 (~100k pixels) which has an ultra-low power mode "down to 36 microW".

https://www.prophesee.ai/event-based-sensor-genx320/

GENX320​

KEY FEATURES
Resolution (px) 320×320
Ultra-low power mode (down to 36μW)
Optical format: 1/5”
Pixel latency @1k lux (μs) <150
Dynamic Range (dB) >120
Nominal contrast threshold (%) 25
Pixel size (μm) 6.3 x 6.3
Embedded features: Anti-flicker filtering (AFK) + Event-rate Controller (ERC) + Spatio-temporal Contrast filter (STC)



The nature of a DVS event sensor is that each pixel only produces an output signal when the change of light impinging on the pixel changes by more than a preset value.

 

[Fullmoonfever]

Surely not already

Can't find details of what the real time accelerators are. Probs not us but....musing if they were an early NDA and been developing the "B" model which will go to market next year I think and if they had developed and tested and prepped for market then would need a licence obviously.

Could also explain the minimal up front licence fee and higher royalty if was a product ready for sale as royalties would be expected sooner

Stranger things have happened I guess haha

#spacecomputing #grmon4 #sparc #riscv #noelv #leon3ft | Frontgrade Gaisler

Our latest GRMON4 debug tool offers enhanced control over SPARC LEON and RISC-V NOEL processors. With advanced features such as advanced breakpoints, GUI, and support for JTAG, Ethernet, and SpaceWire debug connections, it provides a comprehensive debugging experience. A notable addition is...

www.linkedin.com

Frontgrade Gaisler
4,543 followers
21h

Our latest GRMON4 debug tool offers enhanced control over SPARC LEON and RISC-V NOEL processors. With advanced features such as advanced breakpoints, GUI, and support for JTAG, Ethernet, and SpaceWire debug connections, it provides a comprehensive debugging experience. A notable addition is the capability to debug all three processing elements in the new GR716B microcontroller, including the LEON3FT main processor and the two embedded Real Time Accelerators.

Congratulations to Arvid Björkengren and Anton Bäckström for the release of this brand new product! #spacecomputing #grmon4 #sparc #riscv #noelv #leon3ft Explore these new features on our updated website: [https://lnkd.in/dwuH3Pzn)

• 

GR716B | Gaisler

The GR716B is a radiation-hardened microcontroller featuring the fault-tolerant LEON3FT SPARC V8 processor and two Real-Time Accelerators.

www.gaisler.com

Real Time Accelerators
The GR716B includes two real-time accelerators (RTA), whose function is off-loading the main LEON3 processor of demanding real-time control tasks. Each RTA has access to tightly coupled memories for instruction and data, thereby allowing single-cycle access and facilitating the design of realtime application software. This also enables control of multiple real-time critical applications from one and the same RTA.

 

[Diogenese]

Surely not already

Can't find details of what the real time accelerators are. Probs not us but....musing if they were an early NDA and been developing the "B" model which will go to market next year I think and if they had developed and tested and prepped for market then would need a licence obviously.

Could also explain the minimal up front licence fee and higher royalty if was a product ready for sale as royalties would be expected sooner

Stranger things have happened I guess haha

#spacecomputing #grmon4 #sparc #riscv #noelv #leon3ft | Frontgrade Gaisler

Our latest GRMON4 debug tool offers enhanced control over SPARC LEON and RISC-V NOEL processors. With advanced features such as advanced breakpoints, GUI, and support for JTAG, Ethernet, and SpaceWire debug connections, it provides a comprehensive debugging experience. A notable addition is...

www.linkedin.com

Frontgrade Gaisler
4,543 followers
21h

Our latest GRMON4 debug tool offers enhanced control over SPARC LEON and RISC-V NOEL processors. With advanced features such as advanced breakpoints, GUI, and support for JTAG, Ethernet, and SpaceWire debug connections, it provides a comprehensive debugging experience. A notable addition is the capability to debug all three processing elements in the new GR716B microcontroller, including the LEON3FT main processor and the two embedded Real Time Accelerators.

Congratulations to Arvid Björkengren and Anton Bäckström for the release of this brand new product! #spacecomputing #grmon4 #sparc #riscv #noelv #leon3ft Explore these new features on our updated website: [https://lnkd.in/dwuH3Pzn)

• 

GR716B | Gaisler

The GR716B is a radiation-hardened microcontroller featuring the fault-tolerant LEON3FT SPARC V8 processor and two Real-Time Accelerators.

www.gaisler.com

Real Time Accelerators
The GR716B includes two real-time accelerators (RTA), whose function is off-loading the main LEON3 processor of demanding real-time control tasks. Each RTA has access to tightly coupled memories for instruction and data, thereby allowing single-cycle access and facilitating the design of realtime application software. This also enables control of multiple real-time critical applications from one and the same RTA.

Hi FMF,

"Real time control tasks" and "instructions and data" sounds more like a program-controlled processor.

 

[Fullmoonfever]

Hi FMF,

"Real time control tasks" and "instructions and data" sounds more like a program-controlled processor.

Oh well...back in it again it seems.

 

[Earlyrelease]

Oh well...back in it again it seems.

View attachment 74512

Bravo
It’s OK no cats were hurt in this commercial.

 

[Fullmoonfever]

Hi FMF,

"Real time control tasks" and "instructions and data" sounds more like a program-controlled processor.

Hi D

What type of accelerators would fit the description at best guess?

From their docs, no direct info other than the comparison between the older "A" and this "B" shows the RTAs are a new addition at least.

Wonder if Akida could replace those at some point then?

Btw...the cat box was in jest

 

[Diogenese]

Hi D

What type of accelerators would fit the description at best guess?

From their docs, no direct info other than the comparison between the older "A" and this "B" shows the RTAs are a new addition at least.

Wonder if Akida could replace those at some point then?

View attachment 74515

View attachment 74516



Btw...the cat box was in jest

Without digging into it, my first guess would be microcontroller unit (MCU) processors which often have a limited dedicated purpose, but then I decided to do a bit of digging:

https://www.design-reuse-embedded.c...ys-develops-risc-v-core-looks-to-ai-in-space/

Mar. 06, 2024 –

By Nick Flaherty, eeNews Europe (March 1, 2024)

Imsys in Sweden has developed a RISC-V processor core and is part of a project to develop an AI accelerator in space.

Imsys points to a general EU strategy to invest more in the open RISC-V technologyt to strengthen technical independence, stimulate economic growth and increase the pace of innovation, and Imsys is currently applying for EU projects where this is a strength.

RISC-V is also important for designers building chips for space applications, where Imsys works with the Swedish Space Agency and Frontgrade Gaisler and the European Space Agency (ESA).


https://www.frontgrade.com/news/fro...sys-collaborate-integrate-ai-space-technology

Frontgrade Gaisler and Imsys Collaborate to Integrate AI into Space Technology​

20231017
The joint effort by Frontgrade Gaisler and Imsys, both Sweden-based companies, has the potential to make substantial advancements in space technology in two key areas – enhancing AI in space applications and using accelerator technology to optimize system design.

“Working together with the deep competencies at Frontgrade Gaisler, the expert in rugged space applications, is a win-win that will help us provide better solutions and open up new market opportunities -- space, rugged applications, AI, etc..” said Jonas Wӕrn, Chief Executive Officer of Imsys AB. “This is Imsys’ first customer for our new accelerator and of great strategic importance to us. The collaboration will focus on verifying Imsys’ design as well as its ‘on silicon’-construction.”

Because the space industry often faces lengthy development cycles, especially for sophisticated components, adaptable accelerators are essential to keeping pace with AI's rapid evolution. Integrating Imsys’ accelerators into avionics systems would enable more onboard computations, enhancing system responsiveness and optimizing communication efficiency by reducing data transfer between spacecraft and ground stations.

“The opportunity to integrate the Imsys accelerator into our microprocessors aligns with Gaisler’s dedication to advancing space technology,” said Sandi Habinc, General Manager at Frontgrade Gaisler. “Integrating AI computing elements into space avionics systems is a key focus for our future product offerings.”

While many current space systems combine an advanced System on a Chip (SoC) with programmable logic to reduce sensor-to-microprocessor data, the Frontgrade Gaisler and Imsys collaboration aims to eliminate the need for separate programmable logic by successfully integrating hardware accelerator technology into the SoC to streamline the data reduction process and achieve to project’s goal.

Imsys AB is active in the development and design of microprocessors. The company creates solutions for managing information, especially large amounts of data. The company's proprietary and flexible microprocessors have historically been used for highly specialized applications and are now found in AI applications.

##########################

This (RISC-V) would be softwate AI.


https://www.linkedin.com/posts/imsy...laboration-activity-7120064353860513792-pIBI/


Imsys ABImsys AB356 followers356 followers1yr • Edited • 1 year ago
Follow

We are very proud to announce our collaboration with Frontgrade Gaisler to integrate AI accelerators into space technology. This is within the project "New Accelerated Applications in Space".


So in October last year FG and Imsys teamed up to produce RISC-V systems with AI accelerator MPUs.

Within a year or so, FG is flirting with Akida ... "F" is for "fickle".

 

[Fullmoonfever]

Without digging into it, my first guess would be microcontroller unit (MCU) processors which often have a limited dedicated purpose, but then I decided to do a bit of digging:

https://www.design-reuse-embedded.c...ys-develops-risc-v-core-looks-to-ai-in-space/

Mar. 06, 2024 –

By Nick Flaherty, eeNews Europe (March 1, 2024)

Imsys in Sweden has developed a RISC-V processor core and is part of a project to develop an AI accelerator in space.

Imsys points to a general EU strategy to invest more in the open RISC-V technologyt to strengthen technical independence, stimulate economic growth and increase the pace of innovation, and Imsys is currently applying for EU projects where this is a strength.

RISC-V is also important for designers building chips for space applications, where Imsys works with the Swedish Space Agency and Frontgrade Gaisler and the European Space Agency (ESA).


https://www.frontgrade.com/news/fro...sys-collaborate-integrate-ai-space-technology

Frontgrade Gaisler and Imsys Collaborate to Integrate AI into Space Technology​

20231017
The joint effort by Frontgrade Gaisler and Imsys, both Sweden-based companies, has the potential to make substantial advancements in space technology in two key areas – enhancing AI in space applications and using accelerator technology to optimize system design.

“Working together with the deep competencies at Frontgrade Gaisler, the expert in rugged space applications, is a win-win that will help us provide better solutions and open up new market opportunities -- space, rugged applications, AI, etc..” said Jonas Wӕrn, Chief Executive Officer of Imsys AB. “This is Imsys’ first customer for our new accelerator and of great strategic importance to us. The collaboration will focus on verifying Imsys’ design as well as its ‘on silicon’-construction.”

Because the space industry often faces lengthy development cycles, especially for sophisticated components, adaptable accelerators are essential to keeping pace with AI's rapid evolution. Integrating Imsys’ accelerators into avionics systems would enable more onboard computations, enhancing system responsiveness and optimizing communication efficiency by reducing data transfer between spacecraft and ground stations.

“The opportunity to integrate the Imsys accelerator into our microprocessors aligns with Gaisler’s dedication to advancing space technology,” said Sandi Habinc, General Manager at Frontgrade Gaisler. “Integrating AI computing elements into space avionics systems is a key focus for our future product offerings.”

While many current space systems combine an advanced System on a Chip (SoC) with programmable logic to reduce sensor-to-microprocessor data, the Frontgrade Gaisler and Imsys collaboration aims to eliminate the need for separate programmable logic by successfully integrating hardware accelerator technology into the SoC to streamline the data reduction process and achieve to project’s goal.

Imsys AB is active in the development and design of microprocessors. The company creates solutions for managing information, especially large amounts of data. The company's proprietary and flexible microprocessors have historically been used for highly specialized applications and are now found in AI applications.

##########################

This (RISC-V) would be softwate AI.


https://www.linkedin.com/posts/imsy...laboration-activity-7120064353860513792-pIBI/


Imsys ABImsys AB356 followers356 followers1yr • Edited • 1 year ago
Follow

We are very proud to announce our collaboration with Frontgrade Gaisler to integrate AI accelerators into space technology. This is within the project "New Accelerated Applications in Space".


So in October last year FG and Imsys teamed up to produce RISC-V systems with AI accelerator MPUs.

Within a year or so, FG is flirting with Akida ... "F" is for "fickle".

Thanks for getting out the shovel

 

[Diogenese]

Oh well...back in it again it seems.

View attachment 74512

Let meowt!

 

[Earlyrelease]

Without digging into it, my first guess would be microcontroller unit (MCU) processors which often have a limited dedicated purpose, but then I decided to do a bit of digging:

https://www.design-reuse-embedded.c...ys-develops-risc-v-core-looks-to-ai-in-space/

Mar. 06, 2024 –

By Nick Flaherty, eeNews Europe (March 1, 2024)

Imsys in Sweden has developed a RISC-V processor core and is part of a project to develop an AI accelerator in space.

Imsys points to a general EU strategy to invest more in the open RISC-V technologyt to strengthen technical independence, stimulate economic growth and increase the pace of innovation, and Imsys is currently applying for EU projects where this is a strength.

RISC-V is also important for designers building chips for space applications, where Imsys works with the Swedish Space Agency and Frontgrade Gaisler and the European Space Agency (ESA).


https://www.frontgrade.com/news/fro...sys-collaborate-integrate-ai-space-technology

Frontgrade Gaisler and Imsys Collaborate to Integrate AI into Space Technology​

20231017
The joint effort by Frontgrade Gaisler and Imsys, both Sweden-based companies, has the potential to make substantial advancements in space technology in two key areas – enhancing AI in space applications and using accelerator technology to optimize system design.

“Working together with the deep competencies at Frontgrade Gaisler, the expert in rugged space applications, is a win-win that will help us provide better solutions and open up new market opportunities -- space, rugged applications, AI, etc..” said Jonas Wӕrn, Chief Executive Officer of Imsys AB. “This is Imsys’ first customer for our new accelerator and of great strategic importance to us. The collaboration will focus on verifying Imsys’ design as well as its ‘on silicon’-construction.”

Because the space industry often faces lengthy development cycles, especially for sophisticated components, adaptable accelerators are essential to keeping pace with AI's rapid evolution. Integrating Imsys’ accelerators into avionics systems would enable more onboard computations, enhancing system responsiveness and optimizing communication efficiency by reducing data transfer between spacecraft and ground stations.

“The opportunity to integrate the Imsys accelerator into our microprocessors aligns with Gaisler’s dedication to advancing space technology,” said Sandi Habinc, General Manager at Frontgrade Gaisler. “Integrating AI computing elements into space avionics systems is a key focus for our future product offerings.”

While many current space systems combine an advanced System on a Chip (SoC) with programmable logic to reduce sensor-to-microprocessor data, the Frontgrade Gaisler and Imsys collaboration aims to eliminate the need for separate programmable logic by successfully integrating hardware accelerator technology into the SoC to streamline the data reduction process and achieve to project’s goal.

Imsys AB is active in the development and design of microprocessors. The company creates solutions for managing information, especially large amounts of data. The company's proprietary and flexible microprocessors have historically been used for highly specialized applications and are now found in AI applications.

##########################

This (RISC-V) would be softwate AI.


https://www.linkedin.com/posts/imsy...laboration-activity-7120064353860513792-pIBI/


Imsys ABImsys AB356 followers356 followers1yr • Edited • 1 year ago
Follow

We are very proud to announce our collaboration with Frontgrade Gaisler to integrate AI accelerators into space technology. This is within the project "New Accelerated Applications in Space".


So in October last year FG and Imsys teamed up to produce RISC-V systems with AI accelerator MPUs.

Within a year or so, FG is flirting with Akida ... "F" is for "fickle".

Dodgy.

Once again wow.

how do you even find this stuff so quick.

You are an asset to this little group and please thank you and thank you twice for sometimes even putting dolphin language into dumb arse speak that even a low key blue collar worker can sometimes understand.

 

[Boab]

Join BrainChip at CES 2025 – Where Edge AI Takes Center Stage

 

[Boab]

CES 2025

Check back frequently for updates, as we’ll be adding new information throughout CES 2024 and beyond.

brainchip.com

 

[7für7]

2024 has been a year full of ups, downs, and (as always) plenty of discussions here. Thanks to everyone who contributes real value …you’re the reason this forum thrives.

To the usual bashers: I hope Santa brings you some better mood and maybe a few facts instead of constant frustration. That would be some progress, right?

Wishing you all a relaxing holiday season, a great start to the new year, and for 2025, let’s hope BrainChip finally shows what Akida is really capable of. I’m staying optimistic …hopefully, you are too!


Cheers,

Ps: I know, Monday and Tuesday are trading days… but better Sooner than later

Aaaaand…Soon in everyone’s home

 

[DingoBorat]

Ps: I know, Monday and Thursday are trading days… but better Sooner than later

Monday and "Tuesday" dummkopf..

 

[Boab]

CES 2025

Check back frequently for updates, as we’ll be adding new information throughout CES 2024 and beyond.

brainchip.com

Spot the typo?
No biggy.

 

[7für7]

Monday and "Tuesday" dummkopf..

Basher