If you've got $10,000 to invest and are looking for some growth stocks to hold for the long term, splitting it between a couple of top artificial intelligence (AI) stocks can be a good idea. This is a bit of an aggressive strategy, although AI leaders have been leading the market higher for the past few years.
Let's look at two AI stocks to invest in now for the long term.
Image source: Getty Images.
Nvidia
Nvidia (NVDA +1.91%) is the king of AI infrastructure, as its graphics processing units (GPUs) are the primary chips used to power AI workloads. The company currently commands a more than 90% market share in the GPU space, thanks to the wide moat it has established through the ecosystem it has developed around its chips. This starts with its CUDA software platform, which is where most foundational AI code was written and optimized for its chips.
Meanwhile, the company's recent acquisition of SchedMD, which developed an open-source workload manager for AI clusters called Slurm, only helps widen its software moat. Slurm is a software platform that helps determine which GPU is assigned to which tasks and when. By acquiring SchedMD, Nvidia gains control over this important AI orchestration layer, which it can then optimize for its chips and hardware and make for a seamless experience.
NASDAQ: NVDA
Key Data Points
In addition to software, Nvidia has also created an edge with its networking portfolio. Its proprietary NvLink interconnect system can transfer data between its GPUs much faster than standard PCIe connections. It pools GPU memory and boosts data throughput, essentially letting its GPUs act as one powerful unit. This optimizes large language model (LLM) training and discourages vendor mixing and matching.
As AI infrastructure spending continues to ramp up, Nvidia is one of the companies best positioned to benefit. It has seen rapid growth, but with the AI race still just in the early innings, it has a long runway of growth still in front of it.
Alphabet
Alphabet (GOOGL +1.93%) (GOOG +1.91%) is another company that is positioned to be a huge AI winner. Ironically, the reason for this is that it has been able to break free of being completely reliant on Nvidia. Alphabet has designed its own custom AI chips that are optimized to work in its Tensorflow framework. It has been using these chips to run internal workloads for more than a decade, and they are now in their seventh generation. While other companies have developed their own custom AI chips, none are as far along or as battle-tested as Alphabet's Tensor Processing Units (TPUs).
The company's TPUs provide a significant structural cost advantage over both cloud computing competitors and companies developing foundational LLMs, like OpenAI. As a preprogrammed ASIC (application-specific integrated circuit), TPUs offer better price-performance and power efficiency, allowing Alphabet to train its Gemini models and run inference for them more cheaply than other models. It also gives its cloud computing business better margins for customers who choose to use its TPUs.
NASDAQ: GOOGL
Key Data Points
Given Alphabet's better economics, this creates a virtuous circle, as it allows it to pour more money into its chips and AI models, making them better and more attractive to customers. Alphabet also has the advantage of being able to infuse its constantly improving Gemini model throughout its entire product ecosystem, from Google search to its Android smartphone operating system to its worker productivity tools.
At the same time, Alphabet has huge advantages as search and generative AI discovery meld into one. Google is already the default gateway to the internet on most devices, and the company has a long-established history of monetizing search through its vast ad network. It also has decades of search and YouTube video history that cannot be matched.
With cost, distribution, monetization, and data advantages, Alphabet is in a prime position to be a long-term AI winner.
