Tesla Is Looking for ‘Exceptional Ability’ in AI Chips. Does That Make TSLA Stock a Buy Here?

• StockStory Top Pick NVDA +1.37%
• TSLA +1.71%

Tesla (TSLA) CEO Elon Musk seems to have found renewed vigor following the approval of his bumper $1 trillion pay package by the company’s shareholders. So much so that in his latest tweet, the world’s richest man has made the bold claim that his company “expect[s] to build chips at higher volumes ultimately than all other AI chips combined.”

Elaborating further, Musk has set a target of designing a new chip and bringing it to volume production every year. This puts the EV leader in direct competition with its partner, Nvidia (NVDA). Yet, some nuances need to be considered.

More News from Barchart

• Why Investors Are Suddenly Watching Nebius Stock Closely

• Dear BBAI Stock Fans, Mark Your Calendars for December 1

• Palantir Stock Sinks 22% From Its Peak. Should You Buy, Sell, or Hold PLTR?

• Tired of missing midday reversals? The FREE Barchart Brief newsletter keeps you in the know. Sign up now!

Tesla Chips: Use Cases And Beyond

A key component of Musk’s strategy to earn his historical payout is artificial intelligence. In fact, it can be said that it is the centrepiece of his strategy that will drive his ambitions around autonomous vehicles, humanoid robots, and energy.

Notably, the chips that have so far been designed by Tesla and are in operation primarily center around its cars, whereas the GPUs of Nvidia and Advanced Micro Devices (AMD) are more for general-purpose data center use.

The first among the Tesla chips was the HW3 chip introduced in 2019. Designed as a System-on-Chip (SoC), the chip runs onboard vehicle inference for Autopilot / Full Self-Driving (FSD). Board-level compute is typically reported around 72 TOPS (tera operations per second) per chip and 144 TOPS per board, and it is built to automotive quality standards and optimized for low cost and power in the car.

The more recent HW4 (also referred to as the AI4) chip was launched in 2023. It replaces HW3 in newer Model S/X and later Model 3/Y builds, and it comes with higher compute for running larger neural networks, upgraded radar, and a new camera suite. Moreover, built with Samsung's 7-nanometer technology (compared to HW3’s 14-nm), the HW4 provides major gains in energy efficiency, transistor density, and thermal behaviour. All this leads to a 2x-4x better performance than the HW3, which is used across Tesla’s fleet of vehicles.

In its humanoid robots, Optimus, the same cadence for HW3 and HW4 chips are used. The chips assist Optimus in vision perception, motion planning, and control and actuation, giving the robots abilities of 3D understanding of space, object recognition, hand-eye coordination, and balance and foot placement.

Finally, the plans are also afoot for the upcoming AI5 and AI6 chips. Specifically for the AI5, the design is set to be finalized as soon as this month, with volume production targeted in 2026. However, Musk recently cautioned that the A15 would not be largely available until mid-2027.

In the most recent earnings call, Musk informed that the AI5 will be 40 times more powerful than the AI4 chip.

Meanwhile, the timeline for the AI6 is naturally later, with early design already in motion.

Notably, Tesla is using a dual-foundry approach for its manufacturing needs, opting for Samsung and Taiwan Semi (TSM).

Thus, by designing its own chips to better its offerings, Tesla is hugely mitigating the risk that leaves many companies vulnerable to chipmakers. Moreover, since the whole process is in-house, Tesla can customize chips according to its needs and build them for specific use cases. Lastly, by opting for a dual foundry approach, Tesla is also not dependent on one manufacturer, although it may be a costly affair due to higher engineering workloads and verification complexities.

Q3 Reflects Some Weakness But Nothing Alarming in the Financials

The third quarter of 2025 delivered mixed results for Tesla. Top-line figures cleared analyst forecasts, yet profitability contracted for a third consecutive year-over-year period. Consolidated revenue climbed 12% to $28.1 billion compared with the prior-year quarter, propelled primarily by non-automotive segments. Energy generation and storage posted a sharp 44% advance to $3.4 billion, while services revenue rose 25% to $3.5 billion. Automotive sales, the traditional mainstay, managed only a 6% gain to $21.2 billion, aided in part by a final surge tied to the winding down of the $7,500 federal EV tax credit.

Earnings per share declined 31% to $0.50, falling short of the $0.56 consensus projection. Operating cash flow held firm at $6.2 billion, and the balance sheet closed the period with $41.6 billion in cash and equivalents against modest short-term debt of $1.9 billion, reflecting ample liquidity.

On the volume front, deliveries totaled 497,099 vehicles, an increase of 7% year over year, while factory output dipped 5% to 447,450 units. The resulting divergence between production and shipments highlights ongoing fluctuations in end-market demand and deliberate inventory management amid intensifying competition in the electric-vehicle space.

Overall, the TSLA stock is up 3.9% on a YTD basis.

Analyst Opinions on TSLA Stock

Analysts have attributed a rating of “Hold” for the TSLA stock, with a mean target price that has already been surpassed. The high target price of $600 indicates upside potential of about 43.6% from current levels. Out of 41 analysts covering the stock, 14 have a “Strong Buy” rating, two have a “Moderate Buy” rating, 16 have a “Hold” rating, and nine have a “Strong Sell” rating.

On the date of publication, Pathikrit Bose did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. This article was originally published on Barchart.com