Analysts at US bank Morgan Stanley have given their support to electric vehicle maker Tesla’s (TSLA:NASDAQ) supercomputer Dojo, the company's custom-built computer which uses its own proprietary technology.

Tesla first released details of ‘Project Dojo’ in 2021 to solve a common problem related to generative-artificial intelligence (AI) and machine learning (ML) training systems, which was that 'the advancement of hardware is well behind the software’ according to the firm.

Tesla’s solution to this problem was to develop its own innovative hardware structure, piece-by-piece, to provide the ‘ideal architecture’ to run Tesla’s computations.

‘The Dojo hardware was developed to accelerate itself without relying on an outboard device, as each chip is its own supercomputer,’ explains Morgan Stanley.

FASTER AUTO-LABELING PROCESS

‘Tesla estimates its auto-labeling process can be achieved 3.2 times faster than existing tools with Dojo. Additionally, FSD V12 (currently in validation stage, scheduled for release to the public by year end) might drastically reduce/remove the need for labeling, freeing up more of Dojo's capacity for model training/simulations’, say the analysts.

Where do investors stand after Tesla’s 57% 2023 share price rally?

The electric vehicle maker plans to spend more than $1 billion on Dojo during 2023.

Tesla’s shares are up over 5% in pre-market trading to $248 on the Morgan Stanley comments, while year-to-date the shares have rocketed nearly 130%.

The analysts believe Tesla’s share price could rise from $250 to $400 due to the ‘Dojo effect’ and forecast an $178 billion uplift to their 2040 Network Services revenue estimates.

The team also views Dojo as ‘beneficial to margins driven by the higher software attach-rate’.

---