Just before last weekend kicked off, Nvidia (NVDA +0.53%) and SK Hynix (SKHY -7.75%) finished some business, signing the largest memory deal in history. As part of the partnership, SK Hynix's subsidiary, SK Telecom, will build a 2-gigawatt AI cloud data center in Korea using Nvidia's Vera Rubin Platform. Meanwhile, SK Hynix will supply Nvidia with high bandwidth memory (HBM) going forward, while the two companies will work together to co-develop future generations of AI memory.
One of the biggest bottlenecks in the AI infrastructure build-out right now is memory, especially HBM. To reduce latency and improve power efficiency, graphics processing units (GPUs) and other AI chips need to be packaged with HBM. Meanwhile, the need for HBM is only growing as inference increasingly focuses on fast memory access rather than raw compute power.
Image source: The Motley Fool.
However, increasing HBM capacity is a challenge. It requires access to the same EUV (extreme ultraviolet lithography) machines that manufacture GPUs and other advanced logic chips, and the supply for these machines is limited, since the only company that makes them is ASML. At the same time, HBM needs upward of 3 times the wafer capacity of regular DRAM, adding another challenge if one wants to rapidly increase supply. HBM also must be manufactured in massive clean rooms, which can take years to build.
My prediction is that this deal will play an important role in both stocks being big winners over the next few years.
Nvidia: Securing its future
Nvidia has long been the dominant player in the AI infrastructure build-out, and its partnership with SK Hynix will only make it stronger. The company has already locked down the training market, as most foundational AI code was written on its CUDA software platform and optimized for its GPUs, giving it a wide moat in this area.
Locking up a massive amount of HBM supply, meanwhile, will help secure its leadership position, especially as the inference market becomes a larger share of the pie. Nvidia's "acquisition" of Groq and its language processing units (LPUs) is already helping it separate the two phases of inference, enabling faster responses. First, its GPUs (packaged with HBM) handle the prefill phase and understand user prompts, while its LPUs, with SRAM (static random-access memory) built directly onto these chips, handle the decode phase, reducing latency and helping deliver faster answers.
NASDAQ: NVDA
Key Data Points
In an HBM market that is very capacity-constrained, having access to such a large amount of capacity while also being able to offload some work to SRAM-based LPUs is a massive supply-chain advantage. Between its Groq deal and this major HBM supply partnership, Nvidia looks well positioned to remain the dominant player in AI infrastructure and to be a strong player in the inference market going forward. That should be a big growth driver for the stock in the years ahead.
Trading at a forward P/E of just over 15 times analyst estimates for fiscal 2028 (ending January 2028) and given its advantages, this is a top AI stock to buy.
SK Hynix: Adding visibility
SK Hynix's deal with Nvidia is a massive windfall. The Korean company has been Nvidia's biggest supplier of HBM, and this partnership helps cement its status and will be a huge revenue growth driver in the coming years.
NASDAQ: SKHY
Key Data Points
The DRAM maker has been working to establish long-term contracts, and this is a major one that will give its business greater visibility. It also helps tie its fortunes to one of the largest and fastest-growing companies in the world, which helps it de-risk a multiyear capital expenditure expansion. It will likely leave other chipmakers scrambling to secure HBM supply, as Broadcom has already done with Samsung, which will keep overall DRAM capacity tight and prices high for the foreseeable future.
With the stock trading at a forward P/E of under 6, the added visibility that this deal creates could easily lead to some multiple expansion. As such, now is a great time to scoop up shares following this huge multiyear deal.
