Almost two decades after multicore processors were introduced to restart advances in computer speed, multicore programming is still devilishly difficult. But the Cologne, Germany-based startup Silexica has developed software tools that are basically multicore compilers to make things easier. And the company on Thursday raised $18 million from investors to fund additional work on its software.
The company, which has 50 employees worldwide, is tapping into the growing use of multicore processors in modern cars, which require much more computing power than microcontrollers can offer for collision avoidance and infotainment functions. Silexica’s tools were designed to automatically optimize code for multicores, making programming easier and boosting the reliability of the code, as Electronic Design has reported.
Major customers include Tier One automotive suppliers as well as wireless equipment manufacturers. The Japanese supplier Denso is using the company’s software tool, SLX, to convert legacy software that would run inside traditional electronic control units for new multicore platforms. Fujitsu has announced that it will incorporate results from Silexica’s software tools into the software running its next-generation base stations.
The electronics architecture of modern cars is changing. Miles of wires connect more than a hundred electronic control units that control everything from windshield wipers to automatic braking systems in high-end vehicles today. But many companies are considering bundling these control units together, replacing them with a handful of multicore processors that can handle all the same functions simultaneously.
“We created SLX to support software professionals facing the biggest challenge in the industry – programming heterogeneous supercomputers,” said Maximilian Odendahl, chief executive and founder of Silexica, in a statement. He added that the company, which was founded in 2014, would focus on improving its software, which was recently released over the cloud, and build simulation tools to help optimize software for autonomous car systems.
Stockholm, Sweden-based EQT Ventures led the latest funding round, which was also financed by existing investors Merus Capital, Paua Ventures, Seed Fonds Aachen and DSA Invest. The company, which also has offices in the United States and Japan, raised $8 million in its first funding round completed in November 2016. Still unclear is whether the company is profitable or how much revenue it generates.