DeePhi Tech, a China-based deep learning technology developer based on reserarch conducted at Tsinghua and Stanford universities, has raised an eight-digit US dollar amount in a series A round featuring semiconductor producers MediaTek and Xilinx, China Money Network has reported.

Tsinghua Holdings, the tech transfer arm of Tsinghua University, also took part in the round, as did venture capital and private equity firm Sigma Square Capital and existing backers GSR Ventures and Banyan Capital.

DeePhi was founded in 2016 and is working on machine learning technology, in particular deep compression, compiling toolchain and system-level optimisation.

The startup utilises Xilinx’s field-programmable gate arrays and multiprocessor system-on-chips for its inference platforms, and will use the series A capital to further develop its technology.

Nazeem Noordeen, Xilinx’s corporate vice-president of acceleration and IP solutions, said: “Empowered by Xilinx technology and expertise, DeePhi solutions provide breakthroughs in efficiency through techniques like network pruning, compression and use of reduced precision data types.

“Our investment in DeePhi is further evidence of our commitment to provide our customers with industry-leading accelerated computing products.”

– A version of this article first appeared on our sister site Global Corporate Venturing.

China-based deep learning technology developer DeePhi Tech has raised an eight-digit US dollar amount in a series A round featuring semiconductor producers MediaTek and Xilinx, China Money Network has reported.

Tsinghua Holdings, the technology transfer arm of Tsinghua University, also took part in the round, as did venture capital and private equity firm Sigma Square Capital and existing backers GSR Ventures and Banyan Capital.

DeePhi was founded in 2016 by researchers from Tsinghua University and Stanford University, and is working on machine learning technology, in particular deep compression, compiling toolchain and system-level optimisation.

The startup utilises Xilinx’s field-programmable gate arrays and multiprocessor system-on-chips for its inference platforms, and will use the series A capital to further develop its technology.

Nazeem Noordeen, Xilinx’s corporate vice-president of acceleration and IP solutions, said: “Empowered by Xilinx technology and expertise, DeePhi solutions provide breakthroughs in efficiency through techniques like network pruning, compression and use of reduced precision data types.

“Our investment in DeePhi is further evidence of our commitment to provide our customers with industry-leading accelerated computing products.”