Tesla soars 10% after Morgan Stanley says its Dojo supercomputer could add $500 billion to its market cap

• Tesla stock soared 10% on Monday, adding more than $70 billion to its market valuation.
• Tesla jumped after Morgan Stanley upgraded the stock and said its Dojo supercomputer could drive 50% more upside.
• "The more we looked at Dojo, the more we realized the potential for underappreciated value in the stock," Morgan Stanley said.

Tesla stock soared as much as 10% on Monday, adding more than $70 billion to its market value after Morgan Stanley upgraded the stock to "Overweight."

Morgan Stanley's renewed bullishness towards the electric vehicle company is driven by the upside potential in its Dojo supercomputer, which the bank says can add another $500 billion to Tesla's market value.

Analysts led by Adam Jonas called Tesla a "top pick" and raised their price target to $400 from $250, representing potential upside of nearly 50% from current levels.

"The same forces that have driven AWS to reach 70% of Amazon total EBIT can work at Tesla, in our view, opening up new addressable markets that extend well beyond selling vehicles at a fixed price. The catalyst? Dojo, Tesla's customer supercomputing effort in the works for the past 5 years," Adam Jonas said.

He said Dojo should help improve Tesla's full-self driving technology, which requires immense computing power that is well-suited for a supercomputer like Dojo. And a better full-self-driving product should lead to more higher-margin revenues for Tesla, as customers decide to include the add-on FSD software capabilities to their car purchase.

"The more we looked at Dojo, the more we realized the potential for underappreciated value in the stock," Jonas said. "We believe Dojo can represent the next step-change in market perception of Tesla."

He estimated that Dojo could drive $6.5 billion in cost savings for Tesla over the next couple of years as the company increases its computing power with the in-house developed supercomputer.

And importantly, Dojo, which is comprised of thousands of GPU chips from Nvidia, can speed up the development of autonomous driving capabilities. 

According to Tesla, Dojo can shorten the time needed for training workloads to less than a week from more than one month.

"The potential efficiency, speed, and cost benefits compared to current GPUs could allow Tesla to materially reduce their autonomy timeline," Jonas said.

Tesla's Dojo supercomputer could also help the company develop its humanoid robot and have broader implications for the various science projects being worked on by Musk. And that's what makes Tesla's ownership of the supercomputer so special for investors.

"The point we'd make to investors is that Tesla is just one part of a larger area of scientific and commercial interests on which Elon Musk is allocating time, financial resources and talent. We see Tesla as one of the more complex and 'unifying' businesses which can serve as an 'experimental lab' to iterate on advanced technologies with high degree of mission difficulty," Jonas said.

If Tesla reaches Morgan Stanley's $400 price target, it would have a valuation of nearly $1.3 trillion. The stock hit a record high of $414.50 in November 2021.