Nvidia has played a pioneering role in the field of artificial intelligence (AI) over the past four years. The company's graphics processing units (GPUs) have been deployed to train popular large language models (LLMs), such as ChatGPT, and it remains the top seller of AI chips by market share.
The semiconductor bellwether controls more than 80% of the AI chip market. However, Nvidia stock has underperformed the broader market in 2026, gaining just 8% as compared to the 35% jump in the Nasdaq-100 Technology Sector index. This underperformance can be attributed to the growing competition, with rivals such as Intel, Broadcom, and Advanced Micro Devices witnessing strong demand for their server processors, graphics cards, and custom AI chips.
The problem for Nvidia is the emergence of alternative chips capable of handling AI workloads in data centers. However, what's worth noting is that all types of AI processors discussed so far rely on a critical component -- high-bandwidth memory (HBM). This is precisely why I believe that South Korean memory giant SK Hynix (SKHY 11.89%) could become the next Nvidia.
Image source: Getty Images.
SK Hynix dominates the HBM market
HBM helps AI accelerators to quickly access massive data sets thanks to their enormous bandwidth. Additionally, HBM reportedly consumes less power than standard dynamic random-access memory (DRAM) because it is placed close to AI accelerators, resulting in lower heat generation and energy usage.
NASDAQ: SKHY
Key Data Points
These characteristics of HBM help solve a key bottleneck in AI data centers. AI accelerators are known for their high processing speeds, which is why they need to be constantly fed with massive amounts of data to perform seamlessly. So, it is easy to see why HBM demand has been so strong that there is an acute supply crunch.
According to DigiTimes, HBM prices could more than double by next year, jumping to $4 to $5 per gigabyte (GB) from an estimated $2 per GB in the second half of 2026. Not surprisingly, Micron Technology noted in December 2025 that it expects the total addressable market (TAM) for HBM to jump from $35 billion in 2025 to $100 billion in 2028.
The actual growth, however, could be much stronger. That's because the consumption of HBM by custom AI processors alone is expected to grow by 35x between 2024 and 2028, according to Counterpoint Research. All this is great news for SK Hynix.
Counterpoint Research reports that SK Hynix's HBM market share stood at an impressive 58% in the first quarter of 2026. That's well ahead of Samsung and Micron, with each memory manufacturer controlling 21% of this massive market. Importantly, SK Hynix is looking to tighten its grip over this market by boosting manufacturing capacity.
The company raised $26.5 billion from its Nasdaq listing and plans to use the proceeds to build additional HBM capacity, among other things. In fact, SK Hynix noted last month that it is on track to double its wafer production capacity by 2030 to support the rapid build-out of AI data centers. This should allow it to maintain its dominance in HBM chips, which is primarily why it has the potential to become the next Nvidia.
The stock is a no-brainer buy
SK Hynix trades at an incredibly attractive 20 times earnings. The forward earnings multiple of 7 further indicates that buying this AI stock is a no-brainer, given its phenomenal growth. Analysts are predicting a 428% increase in SK Hynix's earnings in 2026, followed by a 42% jump next year.
However, the demand-supply dynamics of HBM clearly indicate that its earnings per share could easily increase at a significantly faster pace. What's more, SK Hynix's dominance of the HBM market can fuel tremendous growth for a long time. That's because the HBM market is projected to grow by a whopping 15x by 2035, driven by investments in AI and high-performance computing (HPC) data centers.
As such, SK Hynix has the potential to go on an Nvidia-like bull run over the next few years, which is why buying this tech stock while it trades at cheap multiples looks like the smart move.
