After NVIDIA, the market has started to reprice HBM. But those who truly understand the supply chain know this: GPUs handle computing power, while HBM feeds the data. As large models grow bigger and training demands increase, the bottleneck for AI development is no longer just chip performance—it’s data transfer speed. This is why HBM is becoming one of the most critical infrastructures in the AI era, and SK Hynix happens to be in this prime position. In the past, the market hyped GPUs, but now it’s starting to focus on supply chain bottlenecks. History shows that every industry upgrade follows the same rule: whoever controls the bottleneck holds the pricing power. In the era of new energy, it was lithium mines. In the AI era, it might just be HBM. This is why more and more capital is reassessing the value of SK Hynix. On July 10, SK Hynix officially launched on BitMart’s U.S. stock trading platform. At the same time, BitMart has also kicked off an AI chip-themed trading event. For those following AI, semiconductors, and the U.S. tech stock trends, this could be a signal worth studying. After all, the real big opportunities often don’t appear when everyone sees them—they emerge when the market just begins to reprice.