Esperanto Reaps $58M to Speed Development of Its AI Chip

Xconomy San Francisco — 

Esperanto Technologies, a startup AI chip developer that has operated mostly below the radar since its founding in 2014, announced this week that it raised $58 million in a Series B fundraising round.

Mountain View, CA-based Esperanto is vying with a host of competitors to create new processors for high-order tasks like machine learning and deep learning, while also minimizing power consumption. The field includes big semiconductor companies like Nvidia and tech giants like Google and Facebook, as well as other startups.

Esperanto plans to use its new influx of capital to produce its first generation of 7-nanometer chips based on the RISC-V Instruction Set Architecture and other open standards. The company has now raised a total of $63 million from investors including hardware and cloud storage company Western Digital, which participated in a previous financing round Esperanto raised, in late 2017. The startup didn’t disclose the investors in the Series B funding, saying only that they included “several strategic and venture capital investors.”

“Next-generation applications in machine learning, AI, and real-time analytics require the highest levels of performance and optimization for these advanced workloads,” Western Digital CTO Martin Fink said in Esperanto’s funding announcement. “The RISC-V platform, and Esperanto solutions, free developers to innovate and optimize for special-purpose computing.”

Working in semi-stealth mode, Esperanto has assembled a staff that includes more than a hundred engineers with expertise that spans artificial intelligence, chip design, processor architecture, software and systems, the company said. It was founded by CEO Dave Ditzel, who was the co-founder and former CEO of x86 microprocessor maker Transmeta, and a former computing technology executive at both Intel and Sun Microsystems.

Ditzel is also an evangelist for the RISC (Reduced Instruction Set Computing) architecture. He and David Patterson, a retired UC Berkeley professor of computer science, co-authored of the paper “The Case for RISC.” Patterson joined Google in 2016 to work on computer design for AI, and last year he was part of a team that won a Turing Award.

Photo copyright crstrbrt. Photo courtesy of Depositphotos.com