According to eetimes, ADI recently acquired Flex Logix, an embedded FPGA and AI IP company, and the 10-year-old company‘s technical assets and technical team will be transferred to ADI.

An ADI spokesperson said “By acquiring Flex Logix, ADI can significantly enhance our digital portfolio and further support our efforts to help customers solve their most challenging problems.” The company declined to disclose the terms of the deal or any further details.

Gregory Bryant, executive vice president and president of ADI‘s business units, said in a LinkedIn post, “I am thrilled to welcome the talented team from Flex Logix to ADI! This team of leaders in [eFPGA] technology joins us as we continue to move forward on our path to leadership at the Intelligent Edge. Flex Logix‘s eFPGA technology, which seamlessly integrates FPGA fabrics into SoCs and ASICs, is one of the key building blocks that enable us to build differentiated platforms and help solve our customers‘ biggest challenges.”

According to Flex Logix‘s website, the company‘s technology assets and technical team have been acquired and existing customers have been addressed.

Flex Logix is described as a reconfigurable computing company providing leading-edge eFPGA, DSP/SDR and AI inference solutions to semiconductor and systems companies.Flex Logix was founded in 2014 by Geoff Tate, CEO, Cheng Wang, CTO, and Dejan Markovic, professor at UCLA. Flex Logix was founded in 2014 by Geoff Tate, CEO, Cheng Wang, CTO, and Dejan Markovic, professor at UCLA, based on work Wang did while pursuing his PhD under Markovic‘s supervision.

Flex Logix engineer Fang-Li Yuan, who has been with the company since its inception, said on LinkedIn, “Surprisingly, our first customer 10 years ago became our last customer, who recently decided to acquire us in order to adopt our technology more broadly across their product line. ”

Flex Logix eFPGAs enable volume FPGA users to integrate FPGAs into their companion SoCs, reducing FPGA cost and power consumption by 5-10x and increasing compute density, which is critical for communications, networking, data centers, microcontrollers and more. Its scalable DSP/SDR/AI is the most efficient, delivering higher inference throughput per square millimeter and per watt.Flex Logix supports process nodes from 180nm to 7nm, with 5nm, 3nm and 18A in development.

Flex Logix has claimed that its InferX X1 is the industry‘s fastest and most efficient AI inference chip, which outperforms NVIDIA‘s Xavier NX on the target detection algorithm YOLOv3. Flex Logix had announced the closing of a $55 million funding round in 2021 led by U.S.-based Secret Bank Capital (Mithril).