We're adding a speculative stock to our Bullpen that can benefit from the AI data center boom
Every weekday, the CNBC Investing Club with Jim Cramer releases the Homestretch — an actionable afternoon update, just in time for the last hour of trading on Wall Street. Stocks were making a comeback in late-afternoon trading on Wednesday following a couple of sessions of declines. Uncertainty around the Strait of Hormuz and a peace deal with Iran remains, but the market is taking comfort in President Donald Trump extending the U.S. ceasefire. It’s worth mentioning that the S & P 500 and Nasdaq were rallying nearly 1% and 1.3%, respectively, despite climbing oil prices. American oil benchmark West Texas Intermediate crude was climbing nearly 4% back to the low-$90s per barrel. With a ceasefire still in place, the market is shifting its focus from the war to earnings and the AI infrastructure capital expenditures cycle, which was driving continued gains in stocks tied to the data center buildout. Alphabetannouncedthe eighth generation of its custom tensor processor units (TPUs) on Wednesday at its Google Cloud Next conference. It’s actually two new chips: the TPU 8t for training and the TPU 8i for inference, or the running of AI models on a daily basis. Together, the company believes they will “deliver scale, efficiency and capabilities across training, serving and agentic workloads.” According to Alphabet’s blog post, the TPU 8t delivers nearly three times the compute performance per pod over the previous generation, making it a “training powerhouse.” The TPU 8i shines in the most latency-sensitive inference workloads because of its memory bandwidth design. Keep in mind that Broadcom collaborates with Google on these TPUs. Both Alphabet and Broadcom are Club names. Google’s success in developing its own hardware and chips is a key advantage over its AI peers. It enabled Google to more efficiently scale compute while maintaining cost control and capex efficiency. Amazon has a big custom chip business too, and a note published by Morgan Stanley late Tuesday highlighted the case for Amazon’s Graviton custom central processing unit (CPU) and Alphabet’s Axion custom- designed CPU. Both CPUs are Arm -based. The analyst believes agentic AI will increase industry CPU spending to $82.5 billion in a base case scenario in 2030 or $110 billion in a bull case scenario, up from about $25 billion today. We view this research as a positive signal not just for Amazon and Alphabet, but also for our newest holding, Arm Holdings, which stands to benefit from a surge in CPU spending over the coming years — both through its licensing and royalty model and its own CPU announced last month. We’re adding a speculative name to our Bullpen as we continue to look for new ideas for the portfolio. The company is called BWX Technologies . It gets most of its sales from the Navy — designing, engineering, and manufacturing nuclear components, reactors, and fuel for nuclear-powered aircraft carriers and subs. BWX also has a growing commercial business connected to civilian nuclear power plants. It designs and manufactures commercial nuclear steam generators, heat exchangers, pressure vessels, reactor components, and other auxiliary equipment. It also supplies large, heavy components to the worldwide nuclear industry. It is the only commercial heavy nuclear component manufacturer in North America. It’s this nuclear power plant angle that has our attention. With AI data centers popping up across the world, the need for more power generation is clear. We see that Wednesday, with the huge rally in GE Vernova shares after a blowout earnings report. BWX is a close partner to GE Vernova. In 2024, it became the first qualified supply chain company to join Vernova’s nuclear business supplier growth. Last year, Vernova awarded a contract to BWX to manufacture the reactor pressure vessel for the first GE Vernova Hitachi Nuclear Energy BWRX-300 small modular reactor to be constructed at Ontario Power Generation’s Darlington nuclear project site. SMRs are more of a long-term story, and BWX stock trades at a rich price-to-earnings multiple of roughly 46 times, explaining why we are calling this a speculative addition to the Bullpen, which acts as our stocks to watch list. Access to power is becoming one of the biggest bottlenecks in the AI industry, and nuclear makes a lot of sense to us as a round-the-clock, clean source of power. It’s an important night of earnings with Tesla , ServiceNow , IBM , Lam Research , Texas Instruments , and United Rentals scheduled to report. Before Thursday’s opening bell, we bet quarterly results from Club names Honeywell and Dover , as well as Dow Inc. , American Express, Keurig Dr Pepper, Lockheed Martin, Comcast, and Blackstone. On the data side, weekly jobless claims and S & P Global PMIs are out in the morning. (See here for a full list of the stocks in Jim Cramer’s Charitable Trust.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
