New memory systems for autonomous driving
In the publicly funded project Memtonomy-2, Fraunhofer IESE is collaborating with project partners from industry and research on new memory systems for autonomous driving.
Last modified:
In the publicly funded project Memtonomy-2, Fraunhofer IESE is collaborating with project partners from industry and research on new memory systems for autonomous driving.
Last modified:
Digitalization enables automation and networking of vehicles, opening up new options for mobility. This is accompanied by a rapidly growing volume of data in vehicles, which must be processed in near-real time. The high demands on computing power, flexibility, and efficiency require new approaches in microelectronics as well as in computing architecture and software architecture.
The partners in the project MEMTONOMY-2 have made it their goal to develop new memory systems for autonomous driving. As part of the joint project, bandwidth is to be increased and energy consumption reduced. This is to be made possible by a new memory controller architecture as well as optimization of the memory data and memory accesses. In addition, safety and reliability are essential when applied in autonomous driving. For in-vehicle applications, memory systems must comply with the safety standard ISO 26262, which is constantly checked and ensured in the project. Due to the high demands and the enormous price pressure, components originally developed for the consumer market are being studied in the project. These include, in particular, memory systems such as DRAM and flash, which play a central role in vehicles due to the constantly increasing amount of data that is generated and needs to be processed.
In the predecessor project MEMTONOMY-1, scientific findings from basic research were processed for transfer to industry and initial solution approaches were already developed. In comparison, the project MEMTONOMY-2 is characterized by a much higher level of industry participation, as it is concerned with demonstrating the solution approaches across the entire value chain – from the memory manufacturer to the memory controller manufacturer and the automotive supplier to the automotive OEM.