News

Xilinx Hails New Single-chip Microprocessors as the “World’s Highest Performing” for ADAS

November 15, 2019 by Lisa Boneta

The single chip integrates 64-bit quad-core Arm Cortex-A53 and the dual-core Arm Cortex-R5 based processing system.

Because most accidents on the road are the result of human error, new automotive technologies are stepping up to increase vehicle and road safety. Some of these technologies include advanced driver-assistance systems (ADAS) and autonomous driving (AD).

Xilinx aims to create more reliable automotive technology with the release of its two newest devices: the Xilinx Automotive (XA) Zynq UltraScale+ MPSoC 7EV and 11EG. 

 

Zynq UltraScale+ MPSoC VIP Architecture

Zynq UltraScale+ MPSoC VIP architecture. Image from Xilinx
 

Based on Xilinx's programmable logic (PL) UltraScale MPSoC architecture, Zynq UltraScale+ MPSoC combines two other dynamic features into a single device: 64-bit quad-core Arm Cortex-A53 and the dual-core Arm Cortex-R5 based processing system (PS).

 

"Highest Performance" Adaptive Devices for Advanced ADAS and AD Applications

Xilinx has a proven track record of forging multi-application hardware; they have spearheaded adaptive and intelligent computing, FPGA, hardware programmable system on a chips (SoC), and the adaptive compute acceleration program (ACAP). 

They claim that the latest parts deliver “the highest programmable capacity, performance and I/O capabilities enabling high-speed, data aggregation, pre-processing, and distribution (DAPD), as well as compute acceleration for L2+ to L4 advanced driver-assistance systems (ADAS) and autonomous driving (AD) applications.”

 

Zynq UltraScale+ EG

Zynq UltraScale+ EG; Zynq UltraScale+ EVEV devices build on the powerful EG platform and are designed for automotive ADAS. Image from Xilinx 

 

They also offer over 650,000 programmable logic cells and nearly 3,000 DSP slices. These changes and improvements are a direct result of customer requests and the pressure to tackle complex challenges in the automotive industry.

 

A Comprehensive Family

The Zynq UltraScale+ MPSoC is a comprehensive device family including single-chip, programmable microprocessors.

 

Zynq UltraScale+ EV

Zynq UltraScale+ EV. Image from Xilinx
 

Additionally, the Xilinx Automotive XA Zynq UltraScale+ MPSoC uses both ARM and FPGA architecture, which enables high system performance, flexibility, scalability, and programmability. Unlike a standard SoC, the MPSoC is a unique addition to Xilinx's SoC portfolio because it enables flexible programming, real-time processing unit, multimedia processing, as well as various peripherals that can be adapted to a wide range of applications.  

 

Meeting the Demands of ADAS

Much of the data that is collected from a sensor must be processed in real-time as to avoid any dangerous or life-threatening accidents. Powerful AI algorithms and image processing play a role to make this a reality.

The Zynq UltraScale+ MPSoC is equipped to handle obstacle detection using a camera with 4+ channels up to 2MPixel and features a dynamic 3D surround. It also includes machine vision object detection for vehicle control and Automatic Emergency Braking (AEB). 

 

ADAS

Zynq UltraScale+ MPSoC can support machine vision for obstacle detection. 


Looking closer at specific data rates, in a white paper on technology and computing requirements for self-driving cars, Intel stated that 1 GB of data needs to be processed every second in the car’s real-time operating system. With the latest UltraScale+ MPSoC, maximum data rates of 6.0 GB/s are supported with minimum rates of 1.5 GB/s.

Competitors, such as TI's TDA2Ex SoC, can deliver maximum speeds of 5.0 GB/s.

 

A Step Closer to Fully Automated Driving  

While automated driving has been tested and implemented in some capacities, there's plenty of room for growth to ensure the safety of fully automated vehicles. Xilinx is providing design examples as well as full-stack solutions with a “variety of networking connectivity options, unique functional safety architecture configurations, and security features.”

Although technology is moving quickly to make safer, high-performing, self-driving cars a reality, legislation will play a major role in determining its ultimate fate.