Micron‘s (NASDAQ: MU) stock has been a huge winner over the past year, as the company has benefited greatly from the ongoing supercyles in the DRAM (dynamic random access memory) and NAND (flash) markets. This has led to explosive revenue growth and ballooning gross margins for the company. This was on full display last quarter, when Micron saw its revenue nearly triple and its gross margin more than double to 74.4%.
However, the company announced perhaps even more important news in mid-March when it revealed that its HBM4 36GB 12-Hi memory, designed specifically for Nvidia‘s Vera Rubin platform, was now in mass production. For graphics processing units (GPUs) and other artificial intelligence (AI) chips to perform their best, they need to be packaged with high-bandwidth memory (HBM). This is because HBM sits next to these chips, allowing them to quickly store, retrieve, and transfer data to speed up processing times.
Will AI create the world’s first trillionaire? Our team just released a report on the one little-known company, called an “Indispensable Monopoly” providing the critical technology Nvidia and Intel both need. Continue »
The move to mass production for the HBM4 is a pivotal moment for Micron. The company has long been considered a technology laggard and more of a fast follower in the memory market compared to Korean companies Samsung and SK Hynix, which were the early leaders in HBM.
However, by getting its HBM4 solution into mass production at the same time as its Korean counterparts, Micron has shown that it is a true competitor ready to grab significant market share in the HBM space moving forward.
Micron’s HBM4 solution is already a strong technological achievement, with more than double the bandwidth of HBM3 and providing a 20% improvement in power efficiency. Given the huge energy costs associated with AI, power improvement efficiencies are always important. Meanwhile, Micron has shown itself to be a leader in this specific area, with its proprietary 1-gamma (1γ) DRAM node.
By designing HBM4 specifically for Nvidia’s Vera Rubin platform, the company is attaching it to perhaps Nvidia’s most important platform. Vera Rubin combines both GPUs and central processing units (CPUs) into one package, and is huge point of emphasis for the chip giant as it looks to transition into being a complete AI infrastructure solution and not just a GPU designer. Meanwhile, CPUs are set to become an increasingly important part of data centers given the rise of agentic AI, as AI agents need more of the orchestration and logic that these chips can provide.
