Home Startups Xilinx Acquires Deep Learning Company DeePhi Tech

Xilinx Acquires Deep Learning Company DeePhi Tech

The competition for accelerators involved in powering machine learning functions such as voice processing and facial recognition appears to be heightening.

In fact, Xilinx, a foremost company in intelligent and adaptive computing, recently revealed that it had bought DeePhi Technology Co. Ltd, which is startup based in Beijing with capabilities in machine learning that specialize in system-level optimization, pruning, and deep compression for neural networks.

The acquisition is meant to help Xilinx in expanding its operations into software tools that operate more efficiently through the building blocks of machine learning through mapping them onto the kind of programmable chips that Xilinx builds. The deal joins up with other machine learning investments at the supplier of fields -programmable gate arrays that are based at San Jose, California.

Xilinx has joined forces with Daimler to create driverless vehicle systems based on its chips, which are typically referred to as FPGAs. What’s more, the company invested over a billion dollars into the creation of its new server chip architecture that targeted both machine learning and big data.

The deal is expected to power the Xilinx’s operations even more. The company is also recruiting an engineering team with unique talent, particularly in FPGAs and machine learning.

READ MORE – 10 Applications of Deep Learning in Business

Furthermore, DeePhi, a company established by researchers drawn from Tsinghua University back in 2016, utilizes tools to not only improve power efficiency but also reduce the memory needs of neural networks that are trained on hundreds of hours of human speech, thousands of pictures, and millions of miles of highway driving.

DeePhi Tech has been creating its machine learning technologies on Xilinx technology platforms. Both companies have worked together since DeePhi Tech’s establishment back in 2016. Even so, its neural network pruning technology has been upgraded to operate on Xilinx FPGAs in a bid to allow impressive performance with remarkable energy efficiency.

Song Yao, DeePhi Tech’s CEO, acknowledged the company’s excitement in continuing its strong collaboration with Xilinx and working even more closely to provide leading machine learning solutions to its customers based not only in China but also other parts of the globe.

Additionally, DeePhi Tech’s CTO Yi Shan said that Xilinx is joining DeePhi in its journey to explore the capability of machine learning and is also backing its innovation as one of the company’s early investors.

Although the financial terms for the acquisition were not revealed, the deal accentuates the growing battle over chips with the capacity to power burgeoning strenuous machine learning functions. In this case, Nvidia is reaping heavily from the software industry‘s move to machine learning.

However, it is spending a lot of billions in a move to repulse industry threats from companies like Google that intend to create custom accelerators to phase out Nvidia’s chips, particularly in data centers.

Other competitors in the field including Graphcore and Wave Computing are some of the startups that have secured over a hundred million dollars for developing custom chips, specifically for neural-network training.

Intel is also spending considerably to continue its dominance over data centers, which is currently being threatened by Nvidia’s shift of focus to accelerators.

Source Xilink

Subscribe to our newsletter

Signup today for free and be the first to get notified on the latest news and insights on artificial intelligence

KC Cheung
KC Cheung has over 18 years experience in the technology industry including media, payments, and software and has a keen interest in artificial intelligence, machine learning, deep learning, neural networks and its applications in business. Over the years he has worked with some of the leading technology companies, building and growing dynamic teams in a fast moving international environment.
- Advertisment -


AI Model Development isn’t the End; it’s the Beginning

AI model development isn’t the end; it’s the beginning. Like children, successful models need continuous nurturing and monitoring throughout their lifecycle. Parenting is exhilarating and, if...