SAN FRANCISCO—Programmable logic vendor Xilinx Inc. Monday (Jan. 31) said it acquired high-level synthesis vendor AutoESL Design Technologies Inc. Financial terms of the deal were not disclosed.
Xilinx (San Jose, Calif.) said expanding its technology foundation and product portfolio to include high-level synthesis would enable the company to serve a broader base of companies where system architects and hardware designers are accustomed to designing at a higher level of abstraction in C, C++ and System C. Most of AutoESL's 25 employees located in Cupertino, Calif. and Beijing will be joining Xilinx, the company said.
Tom Feist, senior marketing director at Xilinx, said the acquisition is part of a strategy by Xilinx to grow the leading programmable logic vendor at the expense of competing products like digital signal processors, where designers are accustomed to designing at a higher level of abstraction.
"There are a finite number of people out there who do VHDL and Verilog," Feist said. "High-level synthesis doesn't completely solve that problem, but you can develop a system with far fewer people who are the RTL guys."
High-level synthesis tools have been used for about 20 years, mostly for ASIC design. But despite the promise of high-level synthesis for improving the efficiency of the design flow, the technology has yet to be adopted on a large scale.
Xilinx plans to make AutoESL's technology available to Xilinx customers as an option to Xilinx' ISE design suite.
Feist said Xilinx has been investing in high-level synthesis companies for more than 10 years. Prior to the acquisition of AutoESL, Xilinx had been an investor in the firm through its technology growth fund, Feist said. He added that Xilinx also invested in other high-level synthesis startups, including Synfora Inc., which was acquired by
Synopsys Inc. last year.
"Over the years we've done a large investment out of the Xilinx venture fund out of the startups," Feist said. "We spent over $10 million in investments trying to nurture these companies to get the technology to the place where it would get the results someone was looking for."
In 2006, Xilinx launched an electronic system level initiative with a goal to help the industry improve quality of results, simplify and abstract design flows, establish interoperability and improve embedded processing flows.
Feist noted that there has been some skepticism that high-level synthesis would ever achieve mass adoption. But, he said, "It's not that we didn't believe in the promise." Instead, Feist said, the question has been about whether the technology was mature enough to address the actual needs of the end user.