The Short-Term View of Tesla’s Future is “Volatile”

It’s been a rather capricious year of car pricing for Tesla. Tesla stock has newly flourished in 2023, up 108% as investors have flocked back to technology stocks and especially AI drivers.

Then again, Tesla stock received 4 downgrades ahead of the release, with analysts expecting vehicle pricing to continue weighing on gross margins. Goldman Sachs lowered expectations for Tesla late this month, citing a difficult pricing environment for EVs. “We’re downgrading shares to neutral from buy, as we believe the stock now better reflects our positive long-term view of the company’s growth positioning,” said the bank’s analysts in a note. Deciding that the market is currently giving Tesla stock more credit for its longer-term opportunities post the recent rally, much consideration for the downgrade was allocated to “the difficult environment for new vehicles that we think will continue to weigh on Tesla’s automotive non-GAAP gross margin this year.”

In April, CEO Elon Musk related to analysts that Tesla is “comfortable” with its 2023 production target of 1.8 million but minimized his previous 4th quarter forecast of a 2 million production number. “These are volatile times,” Musk said. “From a production standpoint, if things go well, we’ve got a shot at 2 million vehicles here. But that is the upside case.”

Looking Ahead to Tesla’s Future Aligned with the Cybertruck

Tesla enthusiasts keep hoping that the company’s Cybertruck deliveries will begin by end of year. Its 2019 announcement has been a bit tempered, however, by German news outlet Handelsblatt’s report that the Cybertruck was initially full of issues concerning the suspension, braking, handling, body sealing, noise levels, and more.

With over 1.5 million pre-orders, eternal optimism may be prevailing over ennui, as Tesla seems to actually be assembling some Cybertrucks. Multitudes of fingers are crossed that the Cybertruck production is methodical and meticulous so that historic company assembly issues don’t follow in a pattern of sputters and delays.

Sam Korus of investment firm ARK Invest thinks the Cybertruck could be a mainstream success that ends up surpassing the Tesla Model Y in popularity. “In ARK’s view, the low expectations for Tesla’s Cybertruck are based on a blind spot,” Korus explains. “Today, most automakers fund the development and sales of unprofitable EVs with profits from their gas-powered truck sales. If the Cybertruck were to disrupt that profit center, traditional automakers could end up in trouble.”

Taking a Longer-Term View

Tesla released its compute capacity forecast this month, drawing a clear line to a trajectory in which it will be not only one of the top EV manufacturers but also one of the most important AI companies in the world.

The projections indicate that by October 2024, the company’s compute capacity will reach 100 exaflops. An exaflop is a measure of performance for a supercomputer that can calculate at least one quintillion floating point operations per second. The Tesla estimate means that it anticipates ~20x more than the 4.5 exaflops it disclosed during AI Day last September. Operating at nearly 100% utilization on its current AI training infrastructure, Tesla is scaling its compute capacity aggressively to capitalize on its significant data advantage. One of the key rationales to do so is win the race to autonomous driving.

If these numbers play out, Tesla may become one of the largest AI training companies in the world, as ARK Investment Management analyzes in a recent newsletter. Tesla intends to scale its capacity from ~14,000 Nvidia A100 chips to the compute equivalent of ~300,000 A100 chips, as shown below. For perspective, with ~80% of the market in the past year, Nvidia shipped an estimated ~350,000 SXM server GPU units.

Tesla Dojo is a supercomputer dedicated to AI machine learning that Tesla launched at AI Day 2021. It uses Tesla-designed chips, its comprehensive infrastructure, and video data from Teslas everywhere to train the neural network. That network is essential support in pursuit of Tesla’s machine vision technology for autonomous driving via its Full Self-Driving system.

ARK projects that, while it probably will continue to purchase hardware from Nvidia, Tesla is likely to shift increasingly to its own hardware — Dojo — to reach 100 exaflops of capacity at ~3× the performance of Nvidia’s A100. Nvidia’s latest chip, called the H100, already is 4–6x more performant than the A100.

With that backdrop, Tesla’s need to accelerate its hardware is keen if it is to dominate the race to full transportation autonomy.

Final Thoughts about Tesla’s Future

Predicting the future is truly a messy business. Some trends point to inevitable systemic change. Other shifts are abrupt and unforeseen. The market capitalization of the company fluctuates around $826 billion. Tesla is once again approaching the symbolic threshold of $1 trillion in market capitalization, which it first reached in October 2021 before falling back.

According to the Musk, the market value and stock market future of Tesla depend on the company’s ability to develop autonomous vehicles. “Really the value of the company is primarily on the basis of autonomy,” Musk said at the Paris Viva Tech innovation conference earlier this month.

Musk’s statement continued a thread Musk began earlier this year with CNBC in which he hinted that Tesla cars could be fully autonomous as early as this year.

Investors need to be continually convinced that Tesla can manufacture safe self-driving cars unlike that of any other car company. Could Tesla’s future rise or fall so easily on one innovation like autonomous driving?