IDTechEx forecasts software-defined vehicles market to be worth more than US$700B by 2034
DOE announces $32.5M in funding for 16 projects to advance transportation electrification

Panasonic Automotive introduces Neuron high-performance compute system (HPC) to advance to software-defined mobility

At CES 2024, Panasonic Automotive Systems Company of America, a tier-one automotive supplier and a division of Panasonic Corporation of North America, announced its High-Performance Compute (HPC) system. Named Neuron, this innovation addresses the rapidly evolving mobility needs anticipated for software-defined vehicle advancements.

As vehicles become more software-reliant, vehicle systems must support the extended software lifecycle by enabling software upgrades and prolonging the supporting hardware capability. Cars rely on hardware and software compute platforms to process, share, sense, and derive insights to handle functions for assisted driving. Panasonic Automotive’s Neuron HPC allows for not only software updates and upgrades but also hardware upgrades across platform lifecycles.

The Neuron HPC can aggregate multiple computing zones to reduce the cost, weight and integration complexity of the vehicle by removing redundant components. Panasonic Automotive’s design supports up-integration with high-performance and heavy data input processing capability. The design is upgradeable, scalable and future-proof across today’s evolving in-vehicle platforms.

Panasonic Automotive's High Performance Compute architecture could reduce the number of distributed electronic control units (ECUs) by up to 80%, allowing for faster, lighter, cross-domain computing for real-time, cross-functional communications. The Neuron HPC design is suited for any mobility platform including internal combustion engine, hybrid, fuel cell or electric vehicles.

With a streamlined design, the Neuron HPC incorporates up-integration capability by consolidating multiple ECUs into one centralized nucleus to handle all levels of ADAS, chassis, body, and in-cabin infotainment features.

Performance features include:

  • Centralized Computing/Distributed Processing: Neuron can integrate various specialized computing processors like GPUs for brilliant visuals and neural processors / accelerators for AI into one centralized unit. Then, that centralized unit can distribute computing across these specialized processors for functional use cases spanning all levels of ADAS, predictive maintenance, safety and security.

  • Scalable & Configurable: Neuron HPC is able to support performance scaled to nearly 10x the most advanced automotive compute core on the road today.

  • Thermal Management: Integrates a liquid cooling design that can dissipate up to 1000W of heat.

  • Speed & Connectivity: The Neuron HPC architecture design also enables upgrading to newer technologies and standards when they are available for automotive use.

  • SDV SW Framework: Neuron provides software framework to support standards based device virtualization, distributed computing and inter-domain communication. It also offers core microservices that can be used in multiple domains.

  • Cyber Security: The Neuron utilizes VERZEUSE, a suite of cyber security technologies and services offered by Panasonic Automotive. It provides security and robustness against cyber-attacks. VERZEUSE enables monitoring of the network data and storage strengthening security-monitoring functions by implementing a multi-tiered verification process against a trusted source.

  • Cloud-Native: The Neuron software platform development is truly “cloud-native.” In November 2023, Panasonic Automotive launched Virtual SkipGen (vSkipGen) – a cloud-based development platform that virtualizes ECU and vehicle development. By activating vSkipGen plus Panasonic Automotive’s industry leading reference cockpit domain controller (CDC) components, this technology enables software updates to drop right onto the Neuron CDC card. This cloud-efficient approach reduces time to market and development spend to enhance the features, efficiencies and design of an HPC system.


The comments to this entry are closed.