BRN Discussion Ongoing

[DingoBorat]

The significance is in the type of AI…….we are not in everything……….yet!…………thats why we have Steve.

This is a part of an email from Tony Dawe and I have to stress, that this is just his opinion, but I think it sums up very well, what is happening.

"Many large tech companies are struggling to adapt to this rapidly changing market, and this is causing significant disruption to the global tech sector"

We can see this clearly IMO, with companies like AMD and their new MI300X Accelerator.

The heat sink for it is so big, they put a bloody handle on it!
Years in development.
Huge performance
Huge power usage.
Huge heat output and a need for stacks of cooling.
Described by a "Tech head" (or possibly even one of their employees) as "a monster".

The Gaudi 3 Accelerator and whatever NVIDIA's latest big power offering is, are in the same boat.

While they were developing these products, the Market has changed around them.

Now the market wants the performance, without the big power usage and everything that stems from that.

Through sheer brilliance and maybe even a dash of dumb luck, BrainChip now has the solutions, the Market craves.

Although I don't think it's his style, Anil Mankar's "like" could have been a bit of a cheeky..
"Hey you know we're here, if you want to talk"

 

[Fullmoonfever]

Whilst this older paper (2022) doesn't mention Akida, I have highlighted one of their statements re neuromorphic AI in autonomous driving which I found encouraging.

I included where the authors worked or were aligned with. Some familiar names there, though Valeo wasn't represented, BMW & Infineon were

Frontiers | Automotive Radar Processing With Spiking Neural Networks: Concepts and Challenges

Frequency-modulated continuous wave radar sensors play an essential role for assisted and autonomous driving as they are robust under all weather and light c...

www.frontiersin.org

ORIGINAL RESEARCH article​

Front. Neurosci., 01 April 2022
Sec. Neuromorphic Engineering
Volume 16 - 2022 | https://doi.org/10.3389/fnins.2022.851774

Automotive Radar Processing With Spiking Neural Networks: Concepts and Challenges​

• 1Chair of Highly-Parallel VLSI-Systems and Neuro-Microelectronics, Faculty of Electrical and Computer Engineering, Institute of Principles of Electrical and Electronic Engineering, Technische Universität Dresden, Dresden, Germany
• 2Infineon Technologies Dresden GmbH & Co., KG, Dresden, Germany
• 3Department of Informatics, Technical University of Munich, Munich, Germany
• 4Infineon Technologies AG, Munich, Germany
• 5BMW Group, Research, New Technologies, Garching, Germany
• 6Centre for Tactile Internet (CeTI) With Human-In-The-Loop, Cluster of Excellence, Technische Universität Dresden, Dresden, Germany

When it comes to AI-based autonomous driving, ensuring functional safety of both software and hardware is a critical issue. The principles that are currently developed to support machine learning models (Henriksson et al., 2018; Mohseni et al., 2019) will also apply to SNNs. Similarly, neuromorphic hardware will have to fulfill the same standards as any automotive electronic system: adhere to temperature ranges, be resistant to vibrations, be deterministic and redundant, or contain self-monitoring. For that reason, only digital neuromorphic systems are candidates for integration in cars, while the use of analog or mixed-signal neuromorphic hardware seems out of scope at the moment due to their intrinsic variability. Hence, we suggest to focus on advanced digital systems such as SpiNNaker2 (Yan et al., 2021) or Loihi2 (Orchard et al., 2021) to further explore neuromorphic hardware for automotive radar processing and automated driving in general

Further to the previous post on the research indicating digital neuromorphic tech is best placed for ADAS, this is an Infineon presso from mid 2022 that also references that research paper. May have been posted before or not but here it is again.

Appears they used Spinnaker2 research chip for their testing but presents some neuromorphic data. Who knows what sort of work since then has been done with Akida.

 

[buena suerte :-)]

Good Luck Buena Suerte

Thanks Worker ...

 

[Iseki]

The best news from CES24 for mine is the 8 customer chip (in concept) designs. Who they are I have no idea.
We might have income from helping with the design work, and hopefully a plan to get these eight in the hands of our ecosystem partners, if needed, who can produce them for the clients at the cheapest possible cost. TBH I'm not sure who in our team is responsible for this getting-it-over-the-line, except Sean himself. Hopefully the momentum of CES24 will push the clients into pulling the trigger on their chips.

 

[David holland]

Has it been 2 yrs since the Mercedes drop, geez along time with no asx excitement

 

[Pom down under]

Has it been 2 yrs since the Mercedes drop, geez along time with no asx excitement

 

[IloveLamp]

Has it been 2 yrs since the Mercedes drop, geez along time with no asx excitement

View attachment 54261

 

[Damo4]

Has it been 2 yrs since the Mercedes drop, geez along time with no asx excitement

Perfect example of a concept taking years to become a product.
Imagine how many (at least 45 companies) who may have concepts that are not yet in a product.
I.e. 100+ NDAs, 55+ named partners/customers/collaborations

Almost as if patience is required

 

[IloveLamp]

Perfect example of a concept taking years to become a product.
Imagine how many (at least 45 companies) who may have concepts that are not yet in a product.
I.e. 100+ NDAs, 55+ named partners/customers/collaborations

Almost as if patience is required

But i don wanna patience

 

[David holland]

Just stating the facts, however I'm stating it must be close as it's been 2 yrs

 

[Iseki]

Has it been 2 yrs since the Mercedes drop, geez along time with no asx excitement

Better question to ask is how many of the other cabin ui/ux suppliers followed on the heels of MB. Maybe some did. We certainly picked up emotion3d as a result. So is there a best design akida chip that can satisfy as many of emotion3d customers as possible? How can we best get this chip manufactured?

 

[Damo4]

But i don wanna patience View attachment 54265

Actual footage of David:

 

[Pom down under]

View attachment 54276

He might be a nice person in real life, but we know what his intentions are.

 

[Bravo]

0% chance……..there is a difference between Cloud/Server AI and Edge AI……….Gaudi 3 is going after NVIDIA Data Centre AI (they throw more power at computation to achieve “AI”…..hence the need for ‘cooling’)………NVIDIA is taking it to the Auto industry and now Intel is following……they think they can use more power to compute the sensor data and use the cloud-to-chip.

Nothing to do with Akida……..just facts……I’m sure everyone on here would prefer we just ignored each….so lets do that.

Intel and Vertiv working on liquid cooling for Gaudi3 AI chip​

---

By Staff Writer, ITWeb
Johannesburg, 12 Dec 2023

---

Vertiv will provide cooling for Intel’s Gaudi3 AI accelerator.

---

Intel is working with digital infrastructure provider Vertiv on a new liquid cooling solution for its Gaudi3 AI accelerator chipset.
Set to launch in 2024, the Gaudi3 processor is designed for the most demanding data centre workloads, AI training, and inferencing, positioning it as Intel's contender against the industry-leading Nvidia AI technology.
The Intel Gaudi3 AI accelerator will enable both liquid-cooled and air-cooled servers, supported by Vertiv pumped two-phase (P2P) cooling infrastructure.
The cooling solution is crucial for AI applications and high-performance computing. Performing intense computations on massive amounts of data leads to increased power consumption and heat generation. Cooling solutions help dissipate this heat, ensuring that the chip operates at its best performance without the risk of overheating.
John Nieman, SVP global thermal line of businesses at Vertiv, says the company is expanding its liquid cooling portfolio to support big tech companies such as Intel.
Liquid cooling also provides a more eco-friendly and cost-efficient solution compared to traditional cooling methods, although it is more complex to install.
“To support increasing thermal design power and heat flux for next-generation accelerators, Intel has worked with Vertiv and other ecosystem
partners

to enable a cooling solution that will help customers meet critical sustainability goals,” says Dr Devdatta Kulkarni, principal engineer of this project at Intel.

The microprocessor giant introduced Gaudi2 in 2022, aiming to rival Nvidia's A100 GPUs as the preferred choice for enterprises training large language models. The acquisition of Israel-based Habana for $2 billion in late 2019 paved the way for this development.

We don't have to ignore each other but, to be honest, I really would prefer that you not accuse me of BS'ing every time I post something.

My point being with Gaudi3 that since Anil Mankar "liked" a post - a post that was very specific to that one particular product, I think increases the chances of us having some involvement at some point in the future. That is my opinion and one that I still stand by from my level of understanding.

For example, we know that Akida is processor agnostic. And, from this previous explanation (below) from Rob Telson on Edge Impulse TAO, I see no reason why Akida IP can't be just as easily incorporated in Intel's GPU's (i.e Gaudi3) in the same way it can NVIDIA's to lower the power consumption and increase the efficiency and performance, which may in turn help to alleviate some of the cooling issues.

As with everything, time will tell.



Rob Telson's response via email to Chapman.

 

[David holland]

View attachment 54276

 

[Damo4]

Further to the previous post on the research indicating digital neuromorphic tech is best placed for ADAS, this is an Infineon presso from mid 2022 that also references that research paper. May have been posted before or not but here it is again.

Appears they used Spinnaker2 research chip for their testing but presents some neuromorphic data. Who knows what sort of work since then has been done with Akida.

View attachment 54256

I don't want to sound too much like a prick, but those who think that dot joining/research hasn't amounted to anything, needs to think about Infineons position. The same logic needs to be applied to every single company testing NM research chips.

There are probably hundreds of not thousands of companies, big and small who have contacted other Vendors offering research chips such as these, including Truenorth, Loihi, Spinnaker etc or even cloud based solutions on Qualcomm and Renesas offerings for example.

Brainchip has spent both time and money on creating platforms that can import models created, or ran, else where. This is also expanded by the now very large ecosystem of vendors and clients they are partnered with.
A good example is the crucial partnership with Arm and Edge Impulse. It's been posted a lot, but Chapman posted a fantastic response about their slides, putting Brainchip im an unbelievably strong position to take on cloud heavy giants like Qualcomm.

When the time comes for one of these companies to pull the trigger, who else is offering Neurotrophic Edge Ai?
Who has also decided they will allow for chips to be used in some cases (VVDN Edge Box)?


All I'm saying is the logic of "no partnership, ecosystem, CES display etc matters unless a dotted like is signed" has to be applied to every single use of an competitors research chips. The potential clients my find themselves in a position to pursue production before these chips are created or licensed.

 

[David holland]

I don't want to sound too much like a prick, but those who think that dot joining/research hasn't amounted to anything, needs to think about Infineons position. The same logic needs to be applied to every single company testing NM research chips.

There are probably hundreds of not thousands of companies, big and small who have contacted other Vendors offering research chips such as these, including Truenorth, Loihi, Spinnaker etc or even cloud based solutions on Qualcomm and Renesas offerings for example.

Brainchip has spent both time and money on creating platforms that can import models created, or ran, else where. This is also expanded by the now very large ecosystem of vendors and clients they are partnered with.
A good example is the crucial partnership with Arm and Edge Impulse. It's been posted a lot, but Chapman posted a fantastic response about their slides, putting Brainchip im an unbelievably strong position to take on cloud heavy giants like Qualcomm.

When the time comes for one of these companies to pull the trigger, who else is offering Neurotrophic Edge Ai?
Who has also decided they will allow for chips to be used in some cases (VVDN Edge Box)?


All I'm saying is the logic of "no partnership, ecosystem, CES display etc matters unless a dotted like is signed" has to be applied to every single use of an competitors research chips. The potential clients my find themselves in a position to pursue production before these chips are created or licensed.

View attachment 54282

Another IP licence around the corner, about to drop folks, I'd be surprised if 1 hasn't dropped around Easter

 

[Slade]

Are you the David Holland that lives in Baconsfield?

 

[Schnitzel lover]

View attachment 54241

CES has come and gone. The AI space is flying. God forbid if it went into a bear market, BRN would be cooked like this turkey for thanksgiving..
Don’t know how lucky you all are.. At least now it may get taken over for a few pennys on the dollar all the hype crew have overpaid to invest in a quacky kwoka..

Close enough..

Sorry mate, you won’t be able to afford Turkey.. It’ll be 2min noodles for the rest of 2024..