SoC Built for IoT and Automotive Edge Processing

March 24, 2023
NXP's i.MX 95 features the eIQ Neutron NPU with up to six Cortex-A55s in the application domain, a M7, and an M33 in two separate lower-power safety domain islands that can be used for various different functions independently of the application domain.

This video appeared in Electronic Design and has been published here with permission.

Check out more of our CES 2023 coverage.

NXP's i.MX95 is a high-performance system-on-chip (SoC) family that targets automotive as well as medical and industrial applications. It includes an array of Arm cores including up to six Cortex-A55s in the application domain plus real-time Cortex-M7 and Cortex-M33 cores. The real-time cores occupy safety-domain islands that can be used for various functions independently of the application domain.

The chip family also incorporates the eIQ Neutron NPU (Fig. 1)The Arm Mali GPU provides display support. I talked with James Prior, Manager of Advanced Edge Processors, about the i.MX95 and its application space. 

The i.MX95 is divided into multiple domains that isolate functionality such as real-time support or security (Fig. 2). The NXP SafeAssure design support allows for creation of solutions that meet ASIL B and IEC61508 compliance. Up to six Arm Cortex-A55 cores make up the application cluster. A Cortex-M7 addresses real-time support while the low-power Arm Corex-M33 core targets safety.

The EdgeLock Secure Enclave hides its own core for managing security functionality and storage. The LPDDR5/4X memory interface handles real-time memory encryption as well as inline memory correction. 

Graphics is based around the Arm Mali GPU. It includes a real-time blend engine and is combined with the NXP image signal processor (ISP) that supports RBG-IR and MIPI-CSI with up to eight virtual channels. The video processing unit (VPU) handles up to 32 video streams. 

Also included in the Flex Domain is the eIQ Neutron neural processing unit (NPU) that accelerates machine-learning (ML) models, including CNN, MLP, RNN, LSTM, and TCN. The eIQ ML Software Development Environment provides tools for taking these models and compiling them for use with the NPU. The hardware performance scales from 32 ops/cycle to 2k ops/cycle. The software provides unified support for the NPU and NXP's range of ML platforms. 

Links

Sponsored Recommendations

Getting Started with Python for VNA Automation

April 19, 2024
The video goes through the steps for starting to use Python and SCPI commands to automate Copper Mountain Technologies VNAs. The process of downloading and installing Python IDC...

Can I Use the VNA Software Without an Instrument?

April 19, 2024
Our VNA software application offers a demo mode feature, which does not require a physical VNA to use. Demo mode is easy to access and allows you to simulate the use of various...

Introduction to Copper Mountain Technologies' Multiport VNA

April 19, 2024
Modern RF applications are constantly evolving and demand increasingly sophisticated test instrumentation, perfect for a multiport VNA.

Automating Vector Network Analyzer Measurements

April 19, 2024
Copper Mountain Technology VNAs can be automated by using either of two interfaces: a COM (also known as ActiveX) interface, or a TCP (Transmission Control Protocol) socket interface...