Tesla Inc.’s (NASDAQ:TSLA) Dojo supercomputer could potentially add up to $500 billion to the company’s market value through the accelerated adoption of robotaxis and network services, according to Morgan Stanley.

Dojo could open up “new markets,” much like AWS did for Amazon Inc (AMZN), wrote analysts Adam Jonas and Daniela M. Haigian in a note, upgrading the stock from “equal” to “above” market valuation. They also raised their 12-month price target from $250 to $400 per share. Tesla has already more than doubled its value this year and closed at $248.50 on Friday.

Tesla shares rose by up to 4% in pre-market trading on Monday.

The supercomputer, designed to handle vast amounts of data in autonomous driving training systems, could give Tesla an “asymmetric advantage” in a potential $10 billion market, these analysts wrote, and it could make software and services the primary driver of value for Tesla in the future.

They also noted the importance of keeping an eye on Tesla’s upcoming full autonomous driving system version expected by year-end and the company’s Artificial Intelligence (AI) Day in early 2024.

Tesla has been mentioning how Dojo gives it an early edge in AI and autonomous driving technology since at least 2021. In July of this year, CEO Elon Musk informed investors that the automaker plans to invest over $1 billion in the project by the end of 2024.

This Morgan Stanley’s baseline target would put the stock close to its all-time high closing of $409.97 in November 2021. This makes the company a notable exception, as the average analyst price target, according to Bloomberg, is $268.42.

“The more we look at Dojo, the more we realize the underappreciated value potential of the shares,” say the Morgan Stanley analysts.