At the SOAFEE Seminar held in Tokyo on May 15, Lenovo launched its next-generation Cloud-Native Automotive Supercomputing Platform (CASP) for vehicle computing. CASP uses the SOAFEE architecture for complete cloud-based development, simulation, and testing of autonomous driving. It allows developers to collaborate in real-time across regions. With flexible computing resources, CASP boosts development efficiency by 300% and cuts costs by 70%, setting a new standard for autonomous vehicle development.
This platform builds on Lenovo’s collaboration with Amazon Web Services (AWS) and is the company’s use of the SOAFEE methodology.
CASP utilizes a hybrid architecture that integrates physical domain controller clusters and a SOAFEE-based cloud-to-edge parity environment. This allows for full cloud-based development, simulation, and testing of algorithms. Developers can work on algorithms and applications in the cloud while connected to remote physical test benches. Using SOAFEE architecture, developers can deploy applications to these test benches with a single click, without needing to recompilation. This ensures consistency between development and real-vehicle deployment, solving the fragmentation issues in traditional automotive software development.
CASP allows developers worldwide to access the cloud in real time for debugging and testing, overcoming geographical and hardware limitations. With flexible computing and automated workflows, it shortens development cycles and boosts efficiency by up to 300%. By using a pay-as-you-go model instead of expensive hardware, it reduces development costs by 70%.
Lenovo Vehicle Computing continues to work closely with SOAFEE members including Arm and AWS, using cloud infrastructure to support efficient development for automakers. This latest upgrade to the CASP platform will make autonomous driving development cheaper and faster.
