The vehicles of our entry-luxury MMA platform will be based on the MB.OS software architecture for the first time in 2024. They will feature many innovations that stem from the tech program behind the Vision EQXX. As part of this, we are refocusing our portfolio and reducing at the same time, complexity. We aim to achieve this with a smaller number of alternatives. Instead of seven, we plan to offer four models in this segment.
Back in 2020, Nvidia had a serious case of vonNeumannitis.![]()
Mercedes-Benz and NVIDIA: Software-Defined Computing Architecture for Automated Driving Across Future Fleet. | Mercedes-Benz Group
Mercedes-Benz, and NVIDIA, the global leader in accelerated computing, plan to enter into a cooperation to create a revolutionary in-vehicle computing system and AI computing infrastructure. Starting in 2024, this will be rolled out across the fleet of next-generation Mercedes-Benz vehicles...group.mercedes-benz.com
Is it safe to say that NVIDIA Hyperion will be using AKIDA enhanced sensors as they will be rolled out onto merc cars by 2024?.... or am I missing something?
I realise this is an old article before everyone goes mental.
I am in awe of your genius. Thankyou for clarifyingBack in 2020, Nvidia had a serious case of vonNeumannitis.
US2020364508A1 USING DECAY PARAMETERS FOR INFERENCING WITH NEURAL NETWORKS
View attachment 12494
A processor, comprising: one or more arithmetic logic units (ALUs) to be configured to identify one or more digital representations of one or more objects based, at least in part, on one or more neural networks trained using one or more decay parameters;
wherein the one or more ALUs are further to be configured to:
apply the one or more decay parameters to prior state information maintained for the one or more neural networks and used for identifying the one or more objects, a weighting of the prior state information being reduced according to the one or more decay parameters;
wherein the one or more ALUs are further to be configured to:
store the state information external to the one or more neural networks and providing the state information to the one or more neural networks for each set of input to the one or more neural networks.
######################################################################
US2022153262A1 OBJECT DETECTION AND COLLISION AVOIDANCE USING A NEURAL NETWORK
View attachment 12495
It may just be possible that MB have knocked a few heads together ...
I have been wondering for two days now who brought MB to Akida and the maybe answer has been here and you knew it all the time but I didn't. I did my own research and ended up validating if not already posted and exactly at the first post in the MB threadBringing across a Mercedes post that I've done on the old forum which shall not be named for newer members to read. There's plenty which could also be brought across so feel free to add anything that you feel is relevant.
__________________
Posted 26/01/2022
I've been doing some latenight reading on all things akida and have stumbled upon some news that I was unaware of and am fairly sure has not been posted on HC before.
Before Mercedes' eventual incorporation of akida into its EQXX voice control system, Mercedes collaborated with Intel to test its neuromorphic processor and its suitability for simple voice commands in vehicles...... Stick that in your pipe and smoke it, Mike Davies!
See the following article by Accenture Labs titled "Driving intelligence at the edge with Neuromorphic Computing".
On page 6 there is a case study titled "Responsive voice control for smart vehicles".
In a recent collaboration with an automotive client, we demonstrated that spiking
neural networks running on a neuromorphic processor can recognize
simple voice commands up to 0.2 seconds faster than a commonly used
embedded GPU accelerator, while using up to a thousand times less
power. This brings truly intelligent, low latency interactions into play, at
the edge, even within the power-limited constraints of a parked vehicle.
The above paragraph left me wondering whether Mercedes was the automotive client in question. I thought the chances were high given that Mercedes is, that I'm aware of, the only automotive business using spiking neural networks within its voice control system.
See page 16 for the Accenture team responsible for the document. Their LinkedIn's can be found here: Alex Kass (Director at Accenture Labs) and Timothy Shea (Research Lead at Intel (previously at Accenture Labs)).
I then did a @ManChildreborn😜 style deep dive on their respective LinkedIn pages and found an interesting post on Tim's LinkedIn. The article linked to his post is titled "Smarter Cars: Auto Makers Experiment With Chips That Think Like Humans".
For context, the article is dated December 10th 2020 and in my view, achieves three things:
1. It confirms, irrefutably, that Mercedes was testing Intel neuromorphic processors as recently as December 2020
2. Peter's Van Der Made's consistent "we have a three-year lead on our competitors" is well founded, at least when it comes to Intel.
3. The article shows how truly well timed and positioned akida is in being the first and only SNN neuromorphic processor available to the market
Support for 1 - It's there in plain english.
Support for 2 - Within the article, and from the horses mouth: "Intel’s neuromorphic chips could begin selling commercially within five years, according to Mike Davies, director of Intel’s Neuromorphic Computing Lab." Five years from December 2020 puts Intel's Loihi commercialisation out to December 2025. Akida was commercially available around December 2021 and so we have a 4 year lead or thereabouts.
Support for 3 - A direct quote from the article: "The chips are expected to be the predominant computing architecture for new, advanced forms of AI deployments by 2025, according to Gartner. By that year, Gartner predicts the technology will displace graphics-processing units." Intel bloody well better hurry up or they won't have a neuromorphic processor on the market when their GPU's are displaced.
Now for some extra fun - if you revert back to the post on Tim's LinkedIn and read the comments, you will find a comment by Alex Kass. Alex refers to the collaboration with the automotive client and provides a link to an "Intel Labs Day 2020: Accenture Session" video that demonstrates Intel's Kapoho Bay / Loihi chip recognising voice commands in a proof of concept system (2:50 - 6:18) such as turning lights on and off, opening doors and starting the engine.
As history shows, Mercedes went with akida in the end.
Why wait until 2025 for a SNN neuromorphic processor when you can have it now?
IMO DYOR
The Electric Software Hub is an epicenter of our research and development and at the same time closely linked to the global production sites. This is where central aspects of the future of Mercedes-Benz become reality - in particular our own operating system MB.OS. Cars are among the most complex products of all. The hardware and software are decoupled and must work together perfectly. We guarantee this in the Electric Software Hub: it is our software integration factory.
The transformation requires even more networked work. With the Electric Software Hub we create the environment for flexibility and creativity. An open room concept, free choice of workplace and various quiet zones indoors and outdoors offer optimal working conditions for our engineers. In this way, the approximately 1,100 colleagues in our new building can develop their full innovative power.
Digitization is the central building block of the automobile production of the future. I am all the more pleased that the Electric Software Hub has now been created here at the Sindelfingen site near Factory 56, our state-of-the-art and fully digitized production facility. Here, the colleagues from development and production take the digitization of the company to a new level together.
The Electric Software Hub will be the technical hub of digitization. With this software integration factory, we are consistently pursuing the path towards e-mobility and shaping the change towards a software-driven mobility provider. The investments that were made here are not only of enormous importance for the future viability of the Sindelfingen location, but also a clear commitment to Germany as a location - and the global crises make it very clear how crucial this is. As employee representatives, we know that the key to success is above all the know-how of the employees. Our experts on the topics of electromobility and software now find the ideal conditions here, to be able to work together across departments and in a state-of-the-art working environment. We are also pleased that around 3,000 software engineers are being hired worldwide - around 1,000 of them at the Sindelfingen site alone. We take the interests of our colleagues seriously and are committed to implementing them together with the company management.
Here is some more info about MB(RDI):
View attachment 14020
View attachment 14021
"How India is shaping Mercedes-Benz's most cutting-edge innovations
Mercedes-Benz Research and Development India (MBRDI) has emerged as the German carmaker's biggest R&D centre globally. It is powering the company's some of the most cutting-edge systems and tech updates
Updated Jun 12, 2022 | 09:24 AM IST
The Stuttgart, Germany-based carmaker Mercedes-Benz’s India research centre is the company’s largest research & development (R&D) facility outside its home country. The Bengaluru-based facility, which draws from the best of India's IT talent pool, is driving the bulk of technology development after over two decades of its establishment as a captive centre. It is here that the majority of the company’s technology development for autonomous vehicles and connected cars occurs in collaboration with global teams.
Mercedes-Benz Research and Development India’s (MBRDI’s) fingerprints are all over the company; it is an indispensable part of everything from infotainment systems to automated solutions and computer vision to the design and production elements of the ultra-premium vehicles.
“We spent our formative years as a student, learning from a company which already had 100 years of car-making history. Now, with the advent of digital, we’ve become the largest R& D hub globally,” said Manu Saale, managing director and CEO, Mercedes-Benz Research and Development India (MBRDI) told Economic Times.
Saale attributed MBRDI’s emergence as a powerhouse of innovation to India’s growing talent pool. It employs some 9000 workers and is expected to hire 1000 more during the year.
MBRDI is in “hyper mode” to recruit people with digital skills, which Saale says is akin to hiring for an IT company. With the break-neck speed of innovation in the car industry, Saale expects high-throttle requirements to continue over the next few years.
Among MBRDI’s many contributions include its work on MBUX (Mercedes-Benz User Experience) which boasts of features like the interior assist—the complete AI software for which was developed at MBRDI. It was also instrumental in developing ‘Hey Mercedes’ feature. The centre is also closely involved in the development of the MB. OS (Mercedes-Benz Operating System) which is expected to be the core software platform for all its cars from 2024."
https://www.mbrdi.co.in/news-room/
You have already seen this info. But I bring it again in full length because from my point of view it is important to recognize the interaction of MB India and Germany and who takes over which work. I assume that they have integrated Akida into the Nvidia system:
Electric Software Hub: Software integration factory at the Sindelfingen site
View attachment 14022
April 08, 2022 - Mercedes-Benz is aiming for a leadership role in both electric driving and vehicle software. In order to accelerate this development, the company has invested more than 200 million euros in the Electric Software Hub in the Mercedes Technology Center (MTC) in Sindelfingen. There, numerous software, hardware, system integration and testing functions are bundled under one roof. At the same time, Mercedes-Benz is intensifying its approach of cross-departmental cooperation.
The Electric Software Hub accelerates the transformation of Mercedes-Benz. A strategic pillar of the company manifests itself here: "Lead in electric drive and car software". From 2025, all new vehicle architectures will be exclusively electric. When it comes to vehicle software, Mercedes-Benz is pursuing a holistic approach that ranges from basic research and development through to the coding of software scopes. In Sindelfingen alone, around 1,000 new jobs are currently being created for software developers. Up to 2,000 additional jobs are currently being added in the global R&D network. The Electric Software Hub brings together the two strategic key issues for the future of Mercedes-Benz and strengthens the role of the Sindelfingen location as a central development and qualification hub.
![]()
Markus Schaefer Member of the Board of Management of Mercedes-Benz Group AG, Chief Technology Officer responsible for development and purchasing
Faster and more efficient development
With increasing digitalization, the control units in the vehicle are taking on more and more complex tasks - from infotainment to driver assistance systems to electric drives. The integration of all components is thus gaining in importance. This process is highly complex. This is all the more true as vehicle development is no longer completed at a certain point. Cars in customer hands also receive continuous software updates with new and improved functions. The Electric Software Hub offers decisive advantages in this new vehicle world: around 1,100 experts from 19 cross-functional departments work closer together than ever. Together with the expansion of digital test procedures, this increases both the speed and the quality of the integration processes. The intensive networking of the disciplines in one building makes the Electric Software Hub exceptional. It stands for Simultaneous Engineering and Seamless Integration at its best.
![]()
With the Electric Software Hub, Mercedes-Benz opens a software integration factory at the Sindelfingen location.
The new building has 70,000 square meters of space spread over eight floors. Inside the Electric Software Hub, the entire electrics/electronics integration process of vehicle development is reflected. The employees ensure that all hardware and software components that are created in research and development work perfectly and interact smoothly. Software and hardware continue to flow into the vehicle from top to bottom – from the code to the product – until they are integrated into vehicle prototypes on the lower surfaces. In this phase, there is an intensive exchange with the global vehicle factories, so that new developments can be optimally transferred to series production.
The most important data of the Mercedes-Benz Electric Software Hub:
![]()
Electric Software Hub Sindelfingen.
Voices on the Electric Software Hub
![]()
Sabine Kohleisen - Member of the Board of Management of Mercedes-Benz Group AG, responsible for Human Resources and Labor Director
![]()
Jörg Burzer - Member of the Board of Management of Mercedes-Benz Group AG, Production and Supply Chain Management
![]()
Ergun Lumali - Deputy Chairman of the Supervisory Board and Chairman of the General Works Council of Mercedes-Benz Group AG"
https://group.mercedes-benz.com/unt...ingen/electric-software-hub-sindelfingen.html
Brilliant dot joining Cosors,
" It (MB India) was also instrumental in developing ‘Hey Mercedes’ feature."
So we can assume that BrainChip India has been working closely with MB India.
Maybe that was the reason we took our Indian programmer contractors in-house? That would give us a higher degree of confidence in the confidentiality of the developments.
ASIC Verification Engineer, sounds like a job for @chapman89Maybe completely unrelated but some BrainChip India dots below - including job description/requirements for advertised position 6-11 months ago
View attachment 14029
View attachment 14030
View attachment 14031
View attachment 14032
View attachment 14033
View attachment 14034
View attachment 14035
View attachment 14036
Notable ways that improved technology “under the hood” will reduce accidents, protect data, and conserve energy include:“Many of the concerns about driverless cars and driver assist systems can be addressed with improved AI/ML operations and internal components,” said BrainChip Founder and CEO Peter van der Made. “Safety is a particularly salient one, but energy efficiency, privacy and security are critical considerations for the automotive industry and their supply chain to address.”
Greener processing“On-chip learning is likely the single biggest breakthrough for the automotive industry, because this alone addresses critical concerns like response times and data privacy,” said van der Made. “A car travelling at 110 km/h does 50 meters in one second. A one-second, or even a fraction of a second delay in receiving data from the cloud could be fatal.”
Mechanical monitoring“Combined with on-chip learning, power-efficient processors will also save vast amounts of energy versus transmitting data via the cloud,” said van der Made. “A back-of-the-envelope calculation suggests a 97% energy savings over current technologies, which not only improves the efficiency of the car, it reduces the data center’s energy use.” Each instance that runs on a remote data center is contributing to its huge carbon footprint.
BrainChip’s Akida™ brings artificial intelligence to the edge in a way that is different from existing technologies. The solution is high-performance, small, ultra-low power and enables a wide array of edge capabilities. The Akida AKD1000 and its intellectual property can be used in applications including Smart Home, Smart Health, Smart City and Smart Transportation. These applications include but are not limited to home automation and remote controls, industrial IoT, robotics, security cameras, sensors, unmanned aircraft, autonomous vehicles, medical instruments, object detection, sound detection, odor and taste detection, gesture control and cybersecurity“There are valid reasons to question the safety and security of self-driving cars, but the largest and most dangerous threat is the processing limitation of present technologies,” said van der Made. “Less capable processors are the real obstacle, especially if corners are cut. It is critical that all data is processed within the car itself, that the data be protected from leak or attack, and that these developments do not come at a greater cost to people or even our planet.”