 Hello everyone, my name is Masai Yoshida, Senior Director of Qualcomm. So today, so I'd like to introduce our insight of the connected compute in BICU. So that's the mean to the how, especially for the 5G awards. So how BICU is moving and also the how Qualcomm is providing the solution. So the please here and if you have any comments, so please post it on the crowd. So first, so I'd like to introduce my bio. So I have started my career from the energy from 1987, almost 35 years ago. And I have the application engineer of the semiconductor, like analog and the digital consumer, LSI and the graphics, BDO, like MPEG. And also the microprocessor and the software, like Windows and the Linux on our SOC. So also I have the work on the multi-core architecture and many of the ecosystem partner, like RTOS, Toron and TNG. And also at GeneVy, it arrives for the Linux. From 2011, so I moved to the Alunesus, actually so that it's major from the NEC to the Alunesus. So then so they have the work on the automotive mainly. And so we have, I have the providing to the RCA based solution to the community, including to the AGO. And also they, I have the work on the autonomous vehicle computing consortium as a board of director. And from this year, so I moved to the Qualcomm and as a senior director of the software product management and continuing to work with the automotive, especially for the ADAS and cockpit and also the connectivity. So the, yeah, so as you know, the automotive industry is rapidly evolving and also the new environmental change, like 5G networks and the V6 is a driving for the automotive electronics. And also the digital cluster, digital cockpit and the ADAS for the, it's a very variable for the old board of the passenger and the driver. And then so the service oriented architecture for the to the cloud. This is also the another segment and the driving force of the automotive. So that's kind of the new consumer. So that access to the car owner and also the driver passenger needs has been a driving force of the, all of the automotive and the ecosystems. Again, so the other 5G is now starting the services in the globally. And then so this is also the driving force or the digital transformation into this infrastructure, so 5G infrastructure enabling much more rich media and those of the services. And those of the on-demand cloud computing is, of course there are very high performance cloud computing either right on the cloud, but it can be available in the each terminals and also the endpoints are like a vehicle. And also it's not 5G is not just a smartphone but many of the devices can be connected through the 5G networks. So that the through IoT will be achievable with that. And also the reliability for the connection and the low latency. This is also another aspect of the technology to enabling to the actual, so the seamlessly connecting to the cloud to the terminals. So that make to the edge AI technology like more edge side mean to the access point and so the operators have been providing the AI technology to deducing to the load of the endpoints and also the cloud computing. So this is also a global standard and then many of the global player is working on the 5G. So then so the today, so the over the 170 operator is already deploying to the 5G networks. And the over the 70 country have been deployed. And also the operator is increasing rapidly because of the investing the 5G networks are deploying. So that's kind of the 5G of course influence on the whole future automobile. So therefore, so this is a very key. And also with the 5G, so how automotive electronics and the ecosystem will moving on. So this is today's key topics. So this 5G transformation is a many influence to the many aspect of the automotive like manufacturing and the sales and the maintenance. And also the services and the actual in-car express and also the transportation. So that means to the, this is a very huge influence for the overall automotive industry. So one of the key is so that how that 5G can be enabling to the connectivity and the compute. Means to the, of course, you know, the automotive, not just automotive, but all of the industrial technologies. Now, very moving to the compute oriented like many of the services and also the data mining and also the autonomous has been rely on the heavily compute technologies. But of course, that kind of a compute is required a very high performance. So then, so that's kind of it can be synergy and the connected with the connectivity with the cloud and the end points and the edge. So this is a 5G is a key technology to how it can be binding to the each application in the end point to the computing platform on the cloud and also the edge AI. So that's a 5G is a driving to that kind of the connection of the computing for the end point applications. So this is one of the purpose of the 5G to enabling AI everywhere. For the automotive, yeah, so the, as you know, the case is one of the key word. Maybe it's used a couple of years, but it's a, that case trend is a continuing like a connected autonomous electrics and the shared. And also the digital cockpit and the ADAS has been evolving continuously like a digital cockpit to move into the connected car. And also the issue consolidation with such a multiple issue can be binding to the single box. And then so the autonomous how it can be a cockpit is a working on. So this is also the key technologies and the trend. And the ADAS, so it's a, yeah, almost a level two ADAS is a mandatory for like a Euro end cap, but it's moving to the level two plus and the level three. And for the future, a level four, level five purely the autonomous driving will be appeared. So that's kind of the market trend has been driven to the new architecture evolution. So the, yeah, as a legacy, so the each ECU like a partner body chassis has been a little bit independent and connected very slow can networks. But that's all the rich services and the application and the safety. So the each domain and also the technology can be more evolved to the hierarchical architecture like a central compute as a plane of the car and some of the domain controller to the maintain each of the domains. And the each node or the directly control to the each sensor and actuator like a motors. So that's kind of the hierarchical architecture will be your main stream of the future vehicles. So then, so this is a migration pass of the ECU architecture. So the currently the current vehicle in production is a gateway based architecture for the each domains. So then, so the gateway are connected each domain like a body and chassis pattern that kind of the information through the gateway that harmonized in each operations. So that's kind of the system is tuned to the each operation. But if the system has been moving to the complicated that each in the correlation is a very tough to manage. So then for the 2020 it's moving to the more domain based architecture to each domain can be your control or the domain control center. Then so that the domain controller are communicated to the more higher speed backbone like a ethernet. So that's kind of the architecture now. Many of the OEM is developing and plan to ship in this 2020. And the 2025, so we are discussing with a multiple customer. So this is the more centralized compute architecture. Like I said, so that's kind of the hierarchical architecture to central compute to the zonal controller and the each ECU and the node to control of the vehicles. So that to reduce into the harness. So that's kind of, it's a zonal controller has been located each location like a front, right, front, left, rear, right, rear, left like that. But so of course, so the each control has been a centralized and then so the center right compute had been managed into the all of our application. So then, so the, that's kind of the architecture are the much more computerized vehicle is a more software oriented. That's mean to the software defined all of the behavior of the vehicle. So therefore that's kind of the central compute has been here how it can be easily building the software. This is the key. And also the how that software can be seamlessly work on the each of the application like ADAS, HMI and also the each domain control like a body part rain chassis. So this is a big challenge, but now we are working on and also that this is the key for the to continue to improve the big architecture for the future. So that's a, this is a current our concept to some of the backplane computing like, you know, the edge server architecture. But of course, this is automotive how it can be aligned. This is a, yeah, currently a deeply discussing within the OEM, the tier one and also the second vendor like us. But that's all that, that's kind of the architecture. Also the need to be corroborate with the crowd. So therefore, so the telematics and also the connectivity is a very key for this architecture. The ADAS, the ADAS is one of the key application for the compute because this is quite heavily depend on the performances. So then, so the, yeah, this is a continue to evolving to the right now. So the level three is almost mandatory, but the level two plus and the level three is a key for next generation. That's the level three plus and also the four or five will be a subject to discuss how move on and that this is a big issue to manage into the complexity of the design of the sensor and also the compute. So that's one of the big challenge to the how our central compute architecture can carry on. So again, so the connectivity driving to the vehicle evolution, but of course, so the key technology is required like with the energy efficiency and the compute and the intelligence. So then, so the, that's kind of the requirement. I have the focus, the multiple technical segment like our 5G and BitsX, digital cockpit and the ADAS and ADD and the car to crowds services oriented architecture. So this is the main trend of the market like OEM intend to. And then, so the how it can be resolved to these issues. So this is the key for the next generation electronic architecture. So for the Qualcomm, so we gonna address in these technologies and we gonna provide into the platform so called the digital chassis. So this is a binding to the these key technologies as a platform. And then so the providing to the solution to the OEMs and the T1 customers. So that's a binding to the car to crowd platform and ADAS and ADD, digital cockpit and the telematics. So as a commonly used for the platform through the multiple OEMs and also the multiple customer. And also, of course, this is the platform to involve into the engaging to the many of the stakeholders like our partner software vendors. So then, so we have the aligning to the our products so called a snapdragon. So this is our solution for the automotive and then to hold each of the technology access and then provide a complete and the end to end solution to the customer. So that's our intentional we are doing right now. So that's kind of the digital chassis have the binding to the many of the technology like 5G and setup with X by high Bluetooth connectivity. And of course compute CPU GPU technologies and also the video or related to multimedia technology including the audio and the ADAS accelerator and the computer vision for the ADAS application. Also the security is a very key for the connected vehicle because there is a many of the cybersecurity risk to coming through to the connected. So that we gonna offer to the very robust and strong security frameworks of the basement of the digital chassis. And also the precise location. So the location services, powering communication and the hills and the tools. That's kind of the tools is a very fundamental but we gonna offer that the very basement of our technologies as a digital chassis. That's kind of the digital chassis of course. So this is a very huge program and the challenge. So we cannot do by ourselves only. So we need to work with the many of the collaboration with the partners including the automaker OEMs and the tier one and the Qualcomm. So because this is the heavily depend on the automaker's intention and those of the tier one's intention but also the Qualcomm have been a participate that cycle to provide into the common platform enabling to automaker and the tier one that kind of the digital chassis based new vehicle in time. So this is our concepts. And yeah, we are working on two builds and the offering to this solution. This so-called this snapdog on digital chassis platform to the our customer and those of our partners. So then so they are little bit explained what our snapdog is. So this is a heterogeneous computing platform to the challenges. So the challenge mean to the how the system can be realized to the these four technology axis like a high performance and the concurrency mean to the, yeah, this is the automotive. So this is not just a server. So it needs to be required very high performance that it's have the concurrency operation to the multiple use cases. So this is the key. So that will our technology can be focused with that and the power efficiency. Of course, this is the electronics and also the EV. So this is the battery operated vehicle. So battery EV and also the even probably hybrid. So the, this is the energy efficiency is a very key because it's a directly affected to the how much mile can be a drive in the single charge on the battery. So therefore, so the power essentially is our main focus for us and the low cost. So this is a, yeah, very good technology, but if it's not available in the all of the second grade of the car, it's meaningless. So therefore, so you have the providing to the scalability and also the to fit each of the grade of the vehicle but also the development cost is a very key. So therefore, so our solution is a pre-designed and provided to the customer to minimize the design cost. And the other time, so this is embedded. So that's meant to the automotive, automotive is a very high-speed moving object means to the, it's need to work on the process in time. Otherwise it's very dangerous and also the meaningless. So therefore, real time is a very key. So this is what's our offer and also what's the definition of the heterogeneous compute platform for the automotives. So yeah, so we are offering to the multiple technologies binding to the, our SNAP programs. So this, for example, for the compute, so the CPU, the very high performance and the high energy efficiency CPU. So we build it, so-called Clio. So this is a use same as a mobile and also the many of the design history is applied for the in production. That means the design maturity is very high. And then also the Arduino is our graphics and the multimedia technologies. So this is a very optimized for the embedded. So that's meant to the multiple instance for the multiple requirement, like a map rendering, HMI rendering, some of the computer vision requirement. That's kind of a multiple requirement can be operated on the Arduino GPU. And also the video processing as a part of the Arduino and also the display technology as well. So this is very optimized to the automotive use case because it is slightly different with the mobile, like, you know, the multiple display support and the high resolution display support. So therefore we have the pretty much tailored to the architecture from the mobile to the automotive. And the hexagon is our data operation and the data processing technology. And it's have the several, bind it to several technology, like a tensor accelerator for the AI and also the vision-based compute and also the audio as well, so audio signal processing as well. So this is a binding to the multiple architecture, but the data to what's kind of the data can be operated. And also the security is one of the key. So the secure processing that our optimized to the automotive use case because, so this is a very highly security is required in the, against the cybersecurity, but also the safety requirement for the automotive. And the connectivity is one of the key because the vehicle has been connected to the cloud, but also the many of the components like a smartphone or even the people or some of the big networks. So that's kind of the connectivity has been tailored to the automotive use case and the binding to the our snapdragon architecture. So then, so these architecture is common to the cockpit and also the ADAS, but so that it's a fit into the multiple performance and functional use case for the to fit into the customer's demand for the systems. So the, yeah, one of the key focus is, so the safety and the security and also the how it can be realized, safety security requirement and also the very rich entertainment application. And also the cloud upgrade and also the very rapid evolution of the software and the very robust and the very tight recovery with the hardware. So that's kind of the two different requirements, how we can achieve to this on the system. So virtualization is one of the key for that. So therefore, so we have the pretty much focusing to how our technology to fit into the virtualization, the multiple operating system environments. So this is the current quick branch of the software solution. So the, yeah, we have the support into the multiple option and the combination of the software. So because many of the, our customer, they have their own preference of the software and those of the, they already built it to the software frameworks on top of that. So therefore we have the offer to the multiple solution. Like, you know, the binding to the QNX hypervisor and also the, with the Linux or Android, the green fields open synergy. So that each of the technology have the, each preference to fit with the preference of the OEMs. And also the Linux based container technology. So that also we have the working on that to how it can be easily deployed to the service on the cloud. So then, so we have the binding to the high level of operating system with the multiple or many of the automotive features already on top of the operating system infrastructure that I explained to the previously. So, but so of course that's kind of the feature set with the expanding and also the, it's up to the customer's use case. So therefore, so we are very open to binding to the customer's demanding, but so the, yeah, we are already working on that. For the, to develop with that kind of the software, so the hardware product, difference hardware product, only the one of the key. So therefore, so we have the, providing to multiple hardware difference for the, for example, the cockpit. So the, we have the dedicated cockpit, compute reference platform on, with our snapdragon technologies. And then so it can be a support to the multiple actual use case on the cockpit. Also the ADAS, we are build it to the reference hardware, also called the snapdragon light platform. And then so this is the hardware reference, but also there we are binding to the multiple partners software stack on top of that and to customer start to the, evaluate to the ADAS application on top of that. So this is our technology, but also there we offer to the more flexibility for the function application performance for the multiple product lineup and also the future. So this is the, we have concept of the SKU. So that's a stock unit. So this is one of the key for the defining to the products. So these technologies to the enabling to the multiple features on the same architecture and once customer more tuned to the application or future headrooms is required. So yeah, we can fit into the SKU to meet with the customer demand. So thank you very much for that today. So I am very short to explain to the our solution and we how it can work with. But of course, so this is just the abstraction of the other solution. And so we need to collaborate with our partner and also the customer. So yeah, again, so that if you have the very interested on the solution, so please keep in touch and continue to work with the ecosystem. So thank you very much.