
Nvidia (NASDAQ: NVDA) was a $360 billion company at the beginning of 2023. It has added $1.8 trillion in value since then, and it’s now the third-largest company in the world behind only Apple and Microsoft.
The heightened interest in artificial intelligence (AI) is the primary driver of that value creation. Nvidia makes the industry’s most powerful graphics processing units (GPUs) for data centers, which developers use to build, train, and deploy their AI models. Those chips drove Nvidia’s data center revenue to more than triple in fiscal 2024 (ended Jan. 28), and the momentum looks set to continue in fiscal 2025.
Nvidia is now using some of its newly acquired wealth to invest in other AI companies, which could hint at where the next wave of AI value is created.
Image source: Nvidia.
Nvidia bought five AI stocks at the end of 2023
Nvidia filed its first-ever 13-F with the Securities and Exchange Commission on Feb. 14, and it publicly revealed new holdings in five different stocks:
SoundHound AI (NASDAQ: SOUN), which develops voice recognition and conversational AI technologies.
Arm Holdings, which designs processors for the world’s largest chip companies. This was Nvidia’s largest investment with a value of $147 million at the end of 2023.
Nano-X Imaging, which is improving patient outcomes by using AI to enhance medical imaging. This stock was Nvidia’s smallest investment with a value of less than $0.4 million at the end of 2023.
Recursion Pharmaceuticals, which is using AI to help with drug discovery.
TuSimple Holdings, which developed an autonomous driving platform for the trucking industry.
SoundHound AI stock has been the best performer of the bunch so far, with a 142% gain in 2024 already. It values Nvidia’s stake at around $8.7 million, which doesn’t sound like much, but SoundHound is only a $1.5 billion company.
So should investors follow Nvidia into the conversational AI specialist?
SoundHound has a growing portfolio of AI products
Most of us are familiar with AI chatbots like ChatGPT, Gemini, and Claude. They were originally designed to ingest text-based prompts,…
..