QUALCOMM / BRN

D

Deleted member 118

Guest
MC has been right all along. Qualcomm must have Akida in their hot little hands.
?
 

Slade

Top 20
  • Like
  • Fire
Reactions: 4 users
D

Deleted member 118

Guest
  • Like
  • Haha
Reactions: 3 users

Slade

Top 20
  • Haha
Reactions: 2 users

M_C

Founding Member
Capture.PNG
 
  • Like
Reactions: 10 users

stuart888

Regular
Below is the Qualcomm Mobile VP, explaining how they implement new innovation strategies like Artificial Intelligence. He states they try to put the newest tech in their "Mobile First" framework, so first the phone, since it scales the most.

He jumps into the low power/AI needs right away. The bottom line for people that don't have time to listen, it is all positive news for Brainchip. Any help they provide Brainchip will be large in scale. That is my take.


 
  • Like
  • Fire
  • Love
Reactions: 23 users

M_C

Founding Member
  • Like
  • Fire
  • Love
Reactions: 18 users

M_C

Founding Member
  • Like
Reactions: 5 users
D

Deleted member 118

Guest

How Qualcomm Technologies is unlocking the power of machine learning [video]​

Meet the researchers bringing AI to the connected intelligent edge.
APR 5, 2022
Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.




Qualcomm-image


Since its founding in 2018, Qualcomm AI Research has developed into a vital resource for the industry, impacting virtually every technology vertical, from wireless to automotive, XR, IoT (Internet of Things), and mobile. Our ever-growing team is based in San Diego, Amsterdam, Seoul, and Markham. I’ve had the pleasure to lead this initiative from the beginning. Even though the scope of our work has scaled, our culture has remained guided by purposeful innovation, passionate execution, and openness. These values ring true across the company.

Purposeful innovation​

Our researchers are working on a wide range of topics, spanning from visionary fundamental research to applied research that solves specific challenges in the market today. Virtually every technology can be optimized with machine learning, which is why we are helping to make AI ubiquitous.

Qualcomm-image


Click to see a larger image.

With the breadth of topics that our team is working on, it’s important to take a system approach to full-stack AI optimization and leverage the expertise in power efficiency Qualcomm Technologies has built over the years. We’ve asked the researchers what they are most excited to be working on and how they see the future of their work. This is how they replied:



Passionate execution​

AI research does not happen in silos but is the result of work being done across teams. Finding the right balance of skills and personalities to produce innovation is key. By bringing together hardware and software experts, researchers, and engineers, the work can enable commercialization at scale. Our team has access to the latest in scientific developments and hardware to work on. The open-door policy means that employees can reach out to each other openly across seniority levels and departments, allowing ideas to be exchanged smoothly. Crucially, researchers are given autonomy around the areas of research they are pursuing. Here’s what our team has to say about successful collaborations in AI research:



Openness​

We publish papers every year at top scientific conferences such as NeurIPS, ICLR, ICML, and CVPR. Our team has also published the code for some of their innovative published papers, so that other researchers can replicate the results and build on top of it. This past year Qualcomm AI Research has grown in expertise with the acquisition of TwentyBN and Reservoir Labs. We are proud to be able to relaunch the TwentyBN computer vision datasets Jester and Something Something under the Qualcomm Developer Network umbrella. The contribution to the machine learning community does not end there. This past year we have also been hard at work on releasing new and improved versions of the AI Model Efficiency Toolkit (AIMET) and recently released an AIMET whitepaperdescribing the quantization techniques that developers can use to successfully implement on-device AI. The work of developers and researchers is also made easier with the AIMET Model Zoo, a collection of pre-trained models and recipes for quantizing 32-bit floating point models to 8-bit integers with little loss in accuracy. Here is what our team has to say about how they see their contribution to the wider AI community:



The benefits of on-device AI​

At Qualcomm AI Research, we are inventing, developing, and enabling the commercialization of power-efficient AI. These technologies are becoming increasingly accessible thanks to Qualcomm Technologies, enabling on-device AI with enhanced privacy, lower latency, and lower cost. Having compact neural network models and efficient silicon means technology can take on more tasks successfully from humans, so that they can focus on the things that matter to them. Our AI research team is central in making this possible. And we are hiring:

https://assets.qualcomm.com/mobile-computing-newsletter-sign-up.html






 
  • Like
  • Fire
Reactions: 7 users
D

Deleted member 118

Guest
 
  • Like
  • Fire
Reactions: 2 users
D

Deleted member 118

Guest

Bringing AI research to wireless communication and sensing [video]​

Our fundamental AI research is fueling wireless innovation by combining the strengths of machine learning with wireless domain expertise.
MAY 26, 2022
Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.




Artificial Intelligence (AI) for wireless is already here, with applications in areas such as mobility management, sensing and localization, smart signaling, and interference management. Recently, Qualcomm Technologies prototyped the AI-enabled air interface and announced the Snapdragon X70 5G modem-RF, the world’s first 5G modem with a dedicated AI processor. These developments are possible thanks to our expertise in both wireless and machine learning (ML) based on more than a decade of foundational research.

Wireless and ML have complementary strengths.

Wireless and ML have complementary strengths.

Click to see a larger image.


It turns out that wireless and ML have complementary strengths, but it takes domain knowledge in both to create optimal wireless solutions. Existing wireless technology enables scalable and interpretable solutions, while ML algorithms perform well for complex tasks and generative processes.
There are numerous facets of wireless technology capabilities that can be enhanced with AI including, but not limited to, power saving, channel estimation, positioning, MIMO detection, environmental sensing, beam management, and optimization. In our webinar “Bringing AI research to wireless communication and sensing,” we zoom in on our fundamental research in the areas of ML for improving communications (i.e., generative modelling and neural augmentation) and ML for enabling RF sensing (i.e., self-supervised and unsupervised learning for positioning).

AI is enhancing wireless communications​

Channel models are central to our wireless design, especially for building and evaluating ML solutions. Classical channel models require cumbersome field measurements, contain hard-coded assumptions about propagation characteristics, and model generic scenarios. Neural channel models, by contrast, can accurately match complex field data distribution, be sampled quickly for prototyping purposes, and built from simple traces. With neural augmentation, our MIMO-GAN (generative adversarial network) solution learns the 3GPP communication channel models precisely, using simple traces achieving mean absolute error of less than -18.7 dB for channel gains and less than 3.6 ns for channel delays.
Communication channels are hard to accurately estimate with all variations in propagation characteristics. Tracking time-varying channels can typically be done with classical Kalman filters, which are interpretable and perform well with arbitrary signal-to-noise ratios and pilot patterns. Their limitations, however, are that the filter parameters vary with Doppler values and a single Kalman filter should not be used for all the Doppler values. Standalone ML solutions also have limitations, such as not generalizing well and not being interpretable. We use neural augmentation again, a concept that we feel provides many beneficial design guidelines, to create the neural-augmented Kalman filter. It outperforms the other two methods by capturing the best of both worlds and generalizes to unseen cases.

Neural augmentation of Kalman filters offers the best of both worlds

Neural augmentation of Kalman filters offers the best of both worlds.

Click to see a larger image.


AI is enabling RF sensing​

When we talk about Qualcomm AI Research’s work in enabling RF sensing, we make the distinction between active positioning (with a communications device) and passive positioning (where access points alone are used to determine the position of a device-less person or object).
Active positioning is especially useful indoors and in other locations without a clear line-of-sight to global navigation satellite system (GNSS). Examples of applications of active positioning with RF sensing include indoor navigation, vehicular navigation, AGV tracking, and asset tracking. This technology has already been demonstrated in real-life scenarios, such as precise positioning for the factory of the future.
Current classical precise positioning methods, such as time difference of arrival (TDOA), do not require labels but are not very accurate in non-line-of-sight conditions and do not use multipath information. Current ML methods, such as RF finger printing (RFFP), are very accurate but require a lot of labels and lack robustness when the environment changes. Our industrial precise positioning method, called Neural RF SLAM, achieves the best of the two technologies — in our experiments with unlabeled channel state information (CSI), it achieves on average 43.4 cm accuracy for 90% of users. Using unlabeled data is key to making this precise positioning technology feasible at scale.

With enough unlabeled CSI samples, we can learn the geometry of the environment without labels.

With enough unlabeled CSI samples, we can learn the geometry of the environment without labels.

Click to see a larger image.


Passive positioning with RF sensing also has a variety of use cases across industries, from touchless control for devices to presence detection and sleep monitoring. Existing technology solutions often do not fare well in real-world scenarios or are unfeasible for deployments at scale. Our solution, called WiCluster, is the first weakly supervised passive positioning technique and addresses the existing deployment challenges. It works in strong non-line-of-sight, across multiple floors, and only requires a few labels. For example, in a work environment with offices and conferences rooms, WiCluster achieved precise positioning with mean errors less than 1.1 m and 2.1 m respectively across these room types, and generally achieved test performance comparable to supervised training with more expensive labeling.



We’re excited to highlight our state-of-the-art research in wireless communication and RF sensing and overcoming the challenges of current methods. This will ultimately lead to better user experiences with improved communications and from our devices better understanding the environment.
 
  • Like
Reactions: 3 users
I’ve posted these slides before so apologies for repeating myself but just reviewing this thread and it’s appropriate to draw the link to Qualcomm.

We are a trusted customer of Valeo. This slide is from their presentation:

1653739528934.jpeg


So working on the theory Qualcomm are supplying the software (along with the others in the diagram) and we are supplying the sensor hardware/brains for all the things Akida offers.

It is also widely reported Stellantis and Qualcomm are partnered up so why wouldn’t they pick the best hardware for their sensors as well. Looks like we can help with all their offerings. This next slide is from Stellantis:

1653739744451.png


1653739838105.png


No definitive proof of course but if Akida’s the AI wonder we believe it is why wouldn’t it be the case!

:)
 
  • Like
  • Fire
Reactions: 11 users

M_C

Founding Member
I’ve posted these slides before so apologies for repeating myself but just reviewing this thread and it’s appropriate to draw the link to Qualcomm.

We are a trusted customer of Valeo. This slide is from their presentation:

View attachment 7984

So working on the theory Qualcomm are supplying the software (along with the others in the diagram) and we are supplying the sensor hardware/brains for all the things Akida offers.

It is also widely reported Stellantis and Qualcomm are partnered up so why wouldn’t they pick the best hardware for their sensors as well. Looks like we can help with all their offerings. This next slide is from Stellantis:

View attachment 7986

View attachment 7987

No definitive proof of course but if Akida’s the AI wonder we believe it is why wouldn’t it be the case!

:)
tenor-10.gif
 
  • Like
  • Haha
  • Fire
Reactions: 5 users
D

Deleted member 118

Guest
 
  • Like
  • Love
  • Fire
Reactions: 4 users

Perhaps

Regular
  • Like
  • Fire
Reactions: 12 users

Bravo

If ARM was an arm, BRN would be its biceps💪!
I’ve posted these slides before so apologies for repeating myself but just reviewing this thread and it’s appropriate to draw the link to Qualcomm.

We are a trusted customer of Valeo. This slide is from their presentation:

View attachment 7984

So working on the theory Qualcomm are supplying the software (along with the others in the diagram) and we are supplying the sensor hardware/brains for all the things Akida offers.

It is also widely reported Stellantis and Qualcomm are partnered up so why wouldn’t they pick the best hardware for their sensors as well. Looks like we can help with all their offerings. This next slide is from Stellantis:

View attachment 7986

View attachment 7987

No definitive proof of course but if Akida’s the AI wonder we believe it is why wouldn’t it be the case!

:)


(y)

 
  • Like
  • Fire
Reactions: 5 users

How Qualcomm Technologies is unlocking the power of machine learning [video]​

Meet the researchers bringing AI to the connected intelligent edge.
APR 5, 2022
Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.




Qualcomm-image


Since its founding in 2018, Qualcomm AI Research has developed into a vital resource for the industry, impacting virtually every technology vertical, from wireless to automotive, XR, IoT (Internet of Things), and mobile. Our ever-growing team is based in San Diego, Amsterdam, Seoul, and Markham. I’ve had the pleasure to lead this initiative from the beginning. Even though the scope of our work has scaled, our culture has remained guided by purposeful innovation, passionate execution, and openness. These values ring true across the company.

Purposeful innovation​

Our researchers are working on a wide range of topics, spanning from visionary fundamental research to applied research that solves specific challenges in the market today. Virtually every technology can be optimized with machine learning, which is why we are helping to make AI ubiquitous.

Qualcomm-image


Click to see a larger image.

With the breadth of topics that our team is working on, it’s important to take a system approach to full-stack AI optimization and leverage the expertise in power efficiency Qualcomm Technologies has built over the years. We’ve asked the researchers what they are most excited to be working on and how they see the future of their work. This is how they replied:



Passionate execution​

AI research does not happen in silos but is the result of work being done across teams. Finding the right balance of skills and personalities to produce innovation is key. By bringing together hardware and software experts, researchers, and engineers, the work can enable commercialization at scale. Our team has access to the latest in scientific developments and hardware to work on. The open-door policy means that employees can reach out to each other openly across seniority levels and departments, allowing ideas to be exchanged smoothly. Crucially, researchers are given autonomy around the areas of research they are pursuing. Here’s what our team has to say about successful collaborations in AI research:



Openness​

We publish papers every year at top scientific conferences such as NeurIPS, ICLR, ICML, and CVPR. Our team has also published the code for some of their innovative published papers, so that other researchers can replicate the results and build on top of it. This past year Qualcomm AI Research has grown in expertise with the acquisition of TwentyBN and Reservoir Labs. We are proud to be able to relaunch the TwentyBN computer vision datasets Jester and Something Something under the Qualcomm Developer Network umbrella. The contribution to the machine learning community does not end there. This past year we have also been hard at work on releasing new and improved versions of the AI Model Efficiency Toolkit (AIMET) and recently released an AIMET whitepaperdescribing the quantization techniques that developers can use to successfully implement on-device AI. The work of developers and researchers is also made easier with the AIMET Model Zoo, a collection of pre-trained models and recipes for quantizing 32-bit floating point models to 8-bit integers with little loss in accuracy. Here is what our team has to say about how they see their contribution to the wider AI community:



The benefits of on-device AI​

At Qualcomm AI Research, we are inventing, developing, and enabling the commercialization of power-efficient AI. These technologies are becoming increasingly accessible thanks to Qualcomm Technologies, enabling on-device AI with enhanced privacy, lower latency, and lower cost. Having compact neural network models and efficient silicon means technology can take on more tasks successfully from humans, so that they can focus on the things that matter to them. Our AI research team is central in making this possible. And we are hiring:

https://assets.qualcomm.com/mobile-computing-newsletter-sign-up.html







After listening to these Qualcomm vids, am I wrong to say that Qualcomm have their own AI research departments looking at developing tech similar to that of Brainchip? And if that’s the case would they be using Akida? Or have I got it all wrong?
I must be missing something…..
Cheers
 
  • Like
  • Thinking
Reactions: 3 users

M_C

Founding Member
Capture14.PNG
Capture13.PNG
Capture13.1.PNG
 
  • Like
  • Fire
  • Love
Reactions: 15 users
D

Deleted member 118

Guest
  • Like
Reactions: 3 users
Top Bottom