By Max A. Cherney SAN FRANCISCO, Dec 2 (Reuters) – Startup Vinci said on Tuesday it has raised $36 million to finance its business of building software that can speed chip and other hardware design by significantly accelerating the simulation of such devices. The Palo Alto, California-based company’s software is entering a crowded market that […]
Science
Software firm Vinci, which speeds up hardware simulation, raises $36 million
Audio By Carbonatix
By Max A. Cherney
SAN FRANCISCO, Dec 2 (Reuters) – Startup Vinci said on Tuesday it has raised $36 million to finance its business of building software that can speed chip and other hardware design by significantly accelerating the simulation of such devices.
The Palo Alto, California-based company’s software is entering a crowded market that has similar artificial intelligence-based simulation products made by established chip software design companies such as Cadence and Synopsys.
Vinci said its simulation software uses an in-house-built AI model to boost the speed of chip simulation.
It will pursue heat simulation first and expand into other areas subsequently. Advanced AI chips generate enormous amounts of heat to the point that the latest systems from Nvidia require liquid cooling for many of the configurations.
Vinci CEO Hardik Kabaria said in an interview with Reuters that its AI-powered software is able to improve simulation speeds without the errors often generated by large language models, described by those in the industry as hallucinations. The software is capable of running simulations at virtually any point along the chip design process, he said.
Kabaria declined to identify current customers but said that Vinci is running pilot programs with several big chip companies and 10 chip companies have benchmarked the software.
The $36 million Series A funding round was led by Xora Innovation, and other investors included Khosla Ventures and Eclipse. The company has raised a total of $46 million so far. Vinci did not discuss its valuation.
The company has about 25 employees and plans to employ a usage-based model to generate revenue.
(Reporting by Max A. Cherney in San Francisco; Editing by Muralikumar Anantharaman)
