Around a week ago, it was reported that Elon Musk was planning to build a massive facility for training advanced AI systems by the fall of 2025, which would contain 100,000 Nvidia (NASDAQ: NVDA) H100 GPUs. This was incredible news for NVDA stock investors, as this large order proved that Nvidia’s H100 chips won’t see a lack of demand as the company gears up to release the more powerful Blackwell GPUs.
But Musk just surprised Nvidia investors again, this time with an even bigger order. The Tesla (NASDAQ: TSLA) CEO said on X that he was looking to buy a staggering 300,000 of Nvidia’s new Blackwell B200 AI GPUs to train xAI by next summer. This proves that Musk still thinks that there’s currently nothing better than Nvidia hardware for AI.
Musk’s Massive Blackwell Order
Musk said that the new B200 AI GPUs will be a huge upgrade to X’s current AI GPU cluster, which is powered by 100,000 previous-gen H100 AI GPUs, and he’s extremely right to say so. In fact, Nvidia’s new Blackwell B200 AI GPU has around 4x the AI training performance of H100, and an incredible 30x the inference performance. However, it’s important to keep in mind that Nvidia’s new Blackwell B200 AI GPU uses more power than H100.
But it seems like Musk, as well as other Nvidia customers like Amazon (NASDAQ: AMZN), Microsoft (NASDAQ: MSFT), and Meta (NASDAQ: META), believe that the insane performance improvements are worth it. Even though we’re looking at 300,000 AI GPUs versus 100,000 AI GPUs, the Blackwell B200 GPUs will have astonishing amounts of AI computing power, and that is something that Nvidia nailed when it first announced Blackwell at GTC 2024 earlier this year. Additionally, now that Nvidia announced the Rubin platform for 2026, we can see its customers, including Musk and his companies, already lining up for the newest tech as well.
What it Means for NVDA Stock
X uses a massive array of Nvidia’s AI GPUs for Grok, Elon Musk’s AI bot and ChatGPT competitor. The AI was developed to provide less straightforward answers, as well as witty and comedic answers compared to other chatbots out there. Basically, Musk is trying to take the “robot” aspect out of the AI bot, and he thinks more Nvidia GPUs will help him do that. It’s estimated that an order of this size could cost Musk between $9 billion and $12 billion, as Nvidia’s CEO, Jensen Huang, said that the Blackwell GPUs cost $30,000 to $40,000 apiece.
While this is great news for Nvidia, as it means more money for the company, the math here is rough, as there could be discounts on really large volumes, and prices could also vary depending on what additional components a customer buys. There’s also the fact that Musk might not go ahead with the plan as described, and might instead go for a smaller order. This is because his xAI start-up recently raised $6 billion in its latest venture round, which means that it wouldn’t be able to fund the entirety of this order with its own resources.
Still, Musk’s comments can be seen as an indicator of the way things are moving in the AI market. During the Computex event in Taiwan, Huang said that by 2026, he expects data centers to start having millions of GPUs connected together, and Nvidia will be the main beneficiary of such a scenario. The global data center market is expected to increase to a valuation of $622 billion by 2030, from $301 billion in 2023, and in March of this year, it was reported that Nvidia had 65% of the data center AI chip market. Therefore, it could definitely capitalize on that expected growth, and if it managed to maintain this share, the company could bring in $404 billion in revenue by 2030.
NVDA Stock Forecast
Investors who want to capitalize on this Musk and Nvidia super-alliance could consider investing in xAI through Cathie Wood’s Ark Venture Fund, which recently disclosed that it has taken a stake in the AI startup. The fund also invests in OpenAI, the industry leader behind ChatGPT, as well as other companies in the industry, such as Figure AI and Shield AI.
But Nvidia investors just can’t let go of the fact that Wood sold Nvidia too early, and missed out on more than $1 billion in returns. This is why investors should just load up on more Nvidia shares, instead of investing in funds. The overall average price target for NVDA stock on Wall Street stands at nearly $1,200, and the consensus rating remains a strong buy, breaking down as 37 buy ratings and three holds. So far, no analyst dares issue a sell rating.
All of these analysts are bullish on NVDA stock for a good reason. Overall worldwide AI chip revenue will grow 26% from $53.4 billion in 2023 to $67.1 billion in 2024, according to a recent report from research firm Gartner. That is expected to double to $119 billion by 2027, and with Nvidia’s massive market share and leadership position in AI chips, no one will benefit from this growth more than the company.
Jensen Huang said Nvidia would design new chips annually, and stressed that companies across a wide range of industries that don’t embrace AI will be left behind, saying that only via AI chips can companies control the “computation inflation” of ever-expanding data needs. With all of this in mind, we think that it’s hard to bet against NVDA stock, and it remains a stock to hold for the long-term. The company keeps innovating and entering new industries that are growing rapidly, so this might actually be the safest stock in the entire market to bet your money on.
Disclaimer
Please visit and read our disclaimer here.
