Investors looking to find a chip stock trading at a discount at the beginning of 2026 don't need to look too far. The reason for this is that the largest semiconductor company in the world, Nvidia (NVDA +2.98%), also has one of the most attractively valued stocks.
Its stock trades at a forward price-to-earnings (P/E) ratio of 24.5 times analyst estimates for fiscal 2027 (ending January 2027) and a price/earnings-to-growth (PEG) ratio of less than 0.7 times (with PEGs below 1 typically viewed as undervalued). Meanwhile, the company grew its revenue by 62% last quarter to $57 billion. Even more impressive is that this was a nearly tenfold increase from the $5.9 billion in revenue it produced in fiscal Q3 of 2023.
Image source: Getty Images.
Nvidia is capitalizing on its strong technological position
Nvidia, of course, is the company powering the current artificial intelligence (AI) infrastructure boom. Its graphics processing units (GPUs) are the preferred chips used to train large language models (LLMs), in large part due to its CUDA software platform, which is where most foundational AI code is written. However, the company has also carved out a strong positioning in networking. Its data center networking portfolio, led by its NVLink interconnect systems, grew its revenue by a whopping 162% last quarter to $8.2 billion.
NASDAQ: NVDA
Key Data Points
Nvidia is positioned to continue to deliver strong revenue growth in the future. While competition from custom AI ASICs (application-specific integrated circuits) is increasing, ASICs are hardwired for specific tasks and thus lack the adaptability of GPUs, which offer the flexibility to be reprogrammed in a fast-shifting tech environment.
Nvidia also recently expanded its software moat through the acquisition of SchedMD, the maker of open-source platform Slurm. Slurm helps hyperscalers (owners of large data centers) schedule AI chip usage, and by now owning SchedMD, it can ensure that its chips are optimized for its platform.
Nvidia is taking on the competition and strengthening its offerings
Meanwhile, outside of Alphabet's tensor processing units (TPUs), most AI ASICs tend to be better suited to AI inference, which is less technically demanding. However, Nvidia has gone out and made a deal to bring on the talent and license the technology from Groq, which has designed a chip specifically for inference. It should eventually bring its chip technology into its CUDA architecture, giving Nvidia a stronger position in inference.
With demand for AI infrastructure showing no signs of letting up, Nvidia looks poised to continue to be one of the biggest winners in the AI space. Throw in a cheap stock price and a wide moat, and this is a stock to own for 2026 and beyond.
