The proliferation of auto electronics is creating more edge processing as well as interconnect and bridging opportunities for chip makers able to deliver low-power, secure and reliable platforms with sufficient thermal management. That proliferation is also shortening product design and deployment cycles as car manufacturers, especially EV makers, add processors for ADAS, infotainment and other car systems.

The booming auto and industrial electronics sectors are prompting chip makers to reorient their designs to meet growing demand. Among them is Lattice Semiconductor, which this week unveiled the latest version of its FPGA family primarily targeting automotive applications.

The Certus-NX FPGA family is aimed at low-power, secure automotive electronics applications designed to extend battery life in EVs while addressing pressing thermal management issues that affect processing performance and reliability.

The latest Nexus iteration is based on FD-SOI process technology, enabling the release of “new products at a faster cadence,” said JuJu Joyce, Lattice Semiconductor’s product marketing manager. Among other enhancements, the FPGA architecture accounts for a 100-fold decrease in soft-error rates, Joyce asserted.

While ADAS chip leaders such as Nvidia focus on central processing of sensor and other data, chipmakers such as Lattice Semiconductor concentrate on proliferating edge co-processing of sensor data. Lattice is “really more on the periphery, closer to the sensors,” said Deepak Boppana, the company’s senior marketing director.

That includes the “bridging and aggregation” schemes used to “talk to all those sensors,” Boppana added in an interview.

The chipmaker based in Hillsboro, Oregon, suspects customers are moving away from conventional processor architectures that are “water-falling” from higher-end data center implementations and even MCUs to more readily available FPGA designs. The new Lattice FPGA is aimed at auto applications such as motor and LED controls, in-vehicle networking and sensor data co-processing in ADAS applications.

“We made a strategic investment [when] we saw this whole edge computing trend,” Boppana said. “Because we started early, we are now in a much better position in terms of really having production” of its FPGAs for automotive applications.

Along with ADAS, the Lattice Semiconductor is also promoting its auto FPGA for early “in-cabin AI” apps such as driver monitoring and human presence detection. Mission-critical applications remain elusive, the chipmaker acknowledges.

Last November, Lattice Semiconductor unveiled a design group intended to give customers access to its FPGA-based design and application stack along with its embedded software development expertise.

Power consumption and space constraints are among the most important design concerns for EV manufacturers. Hence, Lattice Semiconductor pitches its auto FPGA as providing four-fold reduction in battery drain along with a three-fold shrinkage of its chip package. Meanwhile, higher I/O density is said to be double the density of competing chips, enabling compact system designs while delivering what the chip maker claims is 70 percent faster I/O.

The auto FPGA also supports PCI Express and Gigabit Ethernet interfaces.

“There’s been a lot of momentum with [FPGA] design activities,” Boppana noted, with product introduction cycles shrinking from as much as five years to as little as two.

Lattice Semiconductor said it is currently sampling its latest FPGA with leading automotive equipment manufacturers, but declined to identify those customers.

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved