Written by Pradeep Chakraborty

Global Semiconductor Alliance (GSA) Automotive Workshop organized a conference today, in partnership with McKinsey & Co., on the future of automotive computing: cloud vs. edge. GSA Automotive Interest Group is neutral, global collaboration platform for the automotive ecosystem.

With the rise of connected cars and increasing demand for compute and processing power in vehicles, OEMs start to experiment with hybrid applications running in the vehicle and on a cloud backend.

Sandro Grigolli, EMEA Executive Director and VP, GSA, welcomed everyone. The first plans of the GSA will be presented in April 5-6, 2021. People are rethinking the architecture of cars. They are looking at workloads running in and out of the car.

Johannes Deichmann.

Johannes Deichmann, Partner, McKinsey & Co., presented the ACES trends that are shaping the future of auto. ACES trends are shaping the future of automotive, and are enabled by advancements in semiconductors and software. OEMs and suppliers are looking toward different concepts and technological enablers. Also, intelligent communication is within and outside the car. EVs are becoming standard. Adoption of ADAS/AD and connectivity will drastically increase over the next few years.

Vehicle architecture is increasingly becoming centralized and software centric. From software developed for specific hardware, software will be separated from hardware, and more centralized, IT-oriented technology, and architecture patterns. The E/E architecture is evolving toward a centralized architecture. Autonomous driving and connectivity will have significant impact on computing and mobile network needs. Other trends include electrification and shared mobility.

Asif Khan.

Asif Khan, Engagement Manager, McKinsey & Co., presented on changes in mobile network architecture. 5G is expected to enable a range of use cases with varying latency and coverage requirements. It will enable enhanced mobile broadband, massive IoT, and mission-critical control. 5G and distributed network deployment is expected to unlock a variety of automotive use cases. 5G provides very-high reliability targeting 99.999 percent for ultrareliable transmissions. Various automotive use cases will be unlocked, ranging from traffic efficiency to automated driving. CV2X use cases will get mass deployed by 2029.

There are four data center archetypes. These are core/cloud, metro edge, micro edge, and onboard. Mobile architectures are evolving to deploy compute/storage closer to the end users. Growth in demand for colocation data centers, especially, micro edge, outpaces the majority of the data center market. Colocation is growing faster than other segments. Hyperscalers can build and own data centers. Enterprisess will grow primarily to on-premise IoT. There are four major factors driving the need for adoption of edge computing. These are bandwidth constrained networks, need for latency, lack of connectivity, and security and regulation.

Ondrej Burkacky.

Alexander von Falkenhausen, Engagement Manager, McKinsey, presented the automotive use cases and requirements. Advances in connectivity and computing can unlock new use cases. We distinguish among four data center archetypes. Without 5G, most use cases will be calculated in the cloud, excepty safety critical ones. With reliable, full-coverage of 5G, edge will become more relevant, and reduce computing and storage cost on vehicles.

Talking about the evolving value chain, Ondrej Burkacky, Senior Partner, McKinsey, said each stakeholder has various plays on the business chessboard. Various players along the value chain, are trying to expand their footprint. There are risks and opportunities for the semiconductor players. They can expand market demand with specialized silicon sharing and software-centric solutions. There are global supply chain risks.

To see the original article and learn more about the author, Pradeep: https://pradeepstechpoints.wordpress.com/2021/10/28/future-of-automotive-computing-cloud-vs-edge/