MCHP Launches Automotive Gen 4 PCIe® Switches Codeplay Software to Collaborate with Andes

MCHP Launches Automotive Gen 4 PCIe® Switch

High-speed, low-latency connectivity solutions for distributed heterogeneous computing systems are essential elements for next-generation autonomous driving applications. Microchip  Technology Inc. ( Nasdaq : MCHP) announced the launch of the first automotive-qualified 4th generation PCIe® switches on the market. These Switchtec™ PFX, PSX and PAX  switch  solutions provide cutting-edge computing interconnect capabilities for advanced driver assistance systems ( AD AS).

“Our portfolio of automotive-qualified Switchtec Gen 4 switches delivers the lowest latency and high bandwidth required to link  CPU and accelerator building blocks used in ADAS applications,” said Krishna Mallampati, vice president of marketing and applications for Microchip ’s Data Center Solutions business unit  . “We are excited to see the results of years of collaboration with key technology partners and bring these solutions to market.”

Switchtec Gen 4 PCIe switches provide high-speed interconnects to support distributed, real-time, safety-critical data processing in ADAS architectures. PCIe is becoming the preferred computing interconnect solution in the automotive industry for the same reasons it has become prevalent in the data center market. It provides ultra-low latency and low-power bandwidth scalability for CPUs and specialized accelerator devices.

Codeplay Software to collaborate with Andes

Codeplay Software®, an industry leader and pioneer of open standard software tools and services for AI, machine learning , and high performance computing , announces support for the AndesCore™ NX27V IP from Andes Technology Corporation . Andes Technology is a leader in high performance/low power IP and  a founding principal member of RISC -V International. The NX27V is an RV64GC vector processor that supports the RISC-V vector specification up to 512-bit VLEN and SIMD width (or DLEN). It allows SoC designers to create next generation compute acceleration solutions that leverage  AI , ML, and HPC at the edge and in the cloud. Initially, Codeplay will provide support through the AndesCore performance simulator, which provides near cycle accurate information.

Compute acceleration solutions require a new programming model to exploit the full capabilities of available processing power. Incorporating one or more vector processor cores, SoC developers and designers can create applications that leverage single instruction/multiple data (SIMD) heterogeneous architectures. Artificial intelligence and machine learning applications require processing large amounts of vector data for applications such as cloud accelerator cards, neural networks and computer vision algorithms in self-driving cars and visual recognition. Powerful vector processors like the NX27V can rapidly increase performance in processing this data.

C++ oldplay’s Acoran software platform supports the NX27V-based simulator, and the SoC will then provide a broad ecosystem of domain-specific optimized libraries for exascale computing and AI. A key foundation of Acoran is SYCL, an open standard programming model that supports heterogeneous programming based on standard ISO C++.

“The NX27V has been adopted by approximately 10 customer SoC projects for data center accelerators. All of these incorporate multiple instances of our vector processors in cluster-based heterogeneous architectures,” said Dr. Charlie Su, President and CTO of Andes Technology. “This exciting collaboration with Codeplay enables us to bring elegant programming solutions to our customers. We are at the beginning of the next wave of SoCs with Domain Specific Architectures (DSAs) for applications ranging from embedded devices to data center accelerators supporting AI and HPC. The growth potential in this area is huge.”

NXP and Ford Collaborate to Deliver Next-Generation Connected Car Experiences and Expanded Services

NXP Semiconductors N.V. (NASDAQ: NXPI ) today announced a collaboration with Ford Motor Company to deliver enhanced driving experiences, convenience and services to its global fleet, including the 2021 Ford F-150 pickup truck, Mustang Mach-E and Bronco S UV . Ford’s new fully connected vehicle architecture uses NXP’s vehicle networking processors and i.MX 8 series processors to upgrade vehicles, help improve customer lifestyles and simplify the ownership experience.

NXP’s vehicle network processors provide secure in-vehicle networking, enabling gateways to quickly deploy over-the-air (OTA) software updates and new services, while processing and sending deep vehicle data to the cloud to drive continuous product improvements and support data-driven services such as vehicle health management.

NXP i.MX 8 series processors help Ford SYNC ® 4 deliver an advanced multimedia user experience with vivid graphics , enabling in-vehicle productivity, cloud services, enhanced voice recognition and location-based applications.