Wave Computing’s MIPS Processors Power 80% of Today’s ADAS-Enabled Automobiles
Wave’s Datacenter-to-Edge Strategy Delivers Smart, Safe, Secure & Scalable Artificial Intelligence Solutions for Automotive OEMs and Suppliers Globally
Wave Computing, the Silicon Valley company that is accelerating artificial intelligence (AI) from the datacenter to the edge, announced its MIPS processor technology is enabling over 80 percent of vehicles equipped with Advanced Driver Assistance Systems (ADAS) from today’s leading automotive manufacturers. This market leadership is based on the intelligent, safety and security capabilities in Wave’s MIPS architecture, which is purpose-built to comply with rigorous automotive ISO-26262 and ASIL standards. In addition to supporting today’s automotive compliance guidelines, Wave Computing’s AI-enabled, automotive-grade, MIPS roadmap will support current ADAS (L1-L2) solutions and the introduction of conditionally-automated and fully-autonomous vehicles (L2-L5) in the years to come.
“Today’s automotive market is being transformed by autonomy, cloud connectivity and the conversion to electric vehicles. These trends require a robust platform that is scalable, intelligent, safe and secure,” said Art Swift, president of Wave’s MIPS IP business. “Adoption of Wave’s MIPS architecture by the majority of today’s leading automotive manufacturers is not only a testament to our ability to meet current safety and security vehicle requirements, but also deliver a solid roadmap for intelligent in-vehicle, scalable domain controllers and Autonomous Domain Control Units.”
Market research indicates the Automotive Artificial Intelligence Market is expected to exceed more than $10 Billion by 2024 with a CAGR of over 35 percent. The increasingly sophisticated capabilities enabled by cloud computing, big data, machine learning, and mobility is driving the desire for AI to be embedded in growing numbers of autonomous, connected, safe and secure automotive subsystems, including:
- Electric Vehicle Battery Management Systems
- V2X Communications & Telematics Units
- Comfort Systems including Cluster and Infotainment
- Integrated Gbit Ethernet, CAN and LIN Bus Controllers
- Autonomous Camera, Radar & Lidar Vision Subsystems
- Electric Drive Train Position Sensor & Motor Control
Beyond demonstrating efficiency in powering ADAS system Domain Controllers, the performance of the MIPS® architecture has been tested and deployed by Valens Semiconductor, a manufacturer of semiconductor solutions for HDBaseT devices. Valens is using MIPS in its in-vehicle networking products.
“Valens’ HDBaseT Automotive product significantly simplifies in-vehicle connectivity infrastructure, while reducing overall costs,” said Daniel Adler, VP and Head of Automotive Business Unit, Valens. “MIPS’ CPUs provide the necessary efficiency, performance and hardware multi-threading technology to enable high-throughput, secure, data processing embedded systems.”
Wave’s MIPS Technology Delivers a Unique “Class” of Automotive IP
Wave Computing offers two classes of MIPS CPU cores designed to address automotive manufacturers’ unique pain points concerning safety, security and smart processing at the edge. Wave’s range of intelligent MIPS IP cores help carmakers easily integrate CPU technology into today’s connected vehicles to expand existing on-board features such as connectivity, infotainment, navigation and driver safety capabilities, while also offering a roadmap to support rapidly evolving industry requirements.
Wave Computing’s two classes of MIPS automotive IP include:
- Domain Controller Electronic Control Units (ECUs) – This 32-bit, ASIL qualified core offers ISA extensions that support Digital Signal Processing (DSP), Single Instruction Multiple Data (SIMD), Floating Point Units (FPUs) and hardware virtualization (VZ) capabilities to provide scalable, multi-domain security for ECU applications. Wave’s MIPS 32-bit processor core has a complete supporting tools eco-system with AI-optimized libraries for Tensorflow and other popular AI frameworks.
- Autonomous Driving Domain Controllers for Advanced Driver Assistance Systems (ADAS) – This 64-bit, ASIL-qualified platform offers many of the same features as the Domain Controller cores and delivers multi-threading, multi-core, multi-cluster support for Open Source-based, real-time applications. A diverse and flexible product-line designed for applications running Linux and other automotive grade, high-level operating systems, the Autonomous Driving Domain IP series can support several of today’s real-time Autonomous systems with the flexibility to adapt to future Autonomous driving requirements. This 64-bit processor core can also be extended with an AI-IP Toolkit that delivers scalable performance with optimized AI libraries for Tensorflow and other popular AI frameworks.
“We’ve been working closely with Wave’s MIPS team to ensure its Safety Requirements Specification (SRS) meets the stringent prerequisites set by the company’s lead customer,” said Dr. Francesco Rossi, Automotive Safety Solution manager, RESILTECH. “Indeed, the results are extremely detailed and will be of great value to System on Chip vendors.”
In addition to delivering unparalleled performance, the fusion of Wave’s AI dataflow technology with MIPS’ IP architecture offers customers investment protection by delivering a single, linearly scalable and highly efficient AI compute platform that enables the extension of AI capabilities to additional autonomous vehicle functions over time.