What the SK hynix-NVIDIA partnership means for AI chips

NVIDIA and SK hynix have expanded a long-running supplier relationship into a multiyear technology partnership. The move underscores how memory has become a central part of the AI hardware race.

NVIDIA and SK hynix announced the partnership on June 7, 2026, stating plans to co-develop next-generation memory for AI factories and support NVIDIA’s broader AI infrastructure roadmap. The agreement also extends across NVIDIA’s Vera Rubin supercomputers, Vera CPUs, RTX Spark PCs, and Jetson Thor robotics platforms.

“SK hynix has been an extraordinary partner to NVIDIA,” said Jensen Huang, NVIDIA’s founder and CEO. “Together, we will codevelop the next generation of memory for AI factories and support the accelerating global expansion of AI infrastructure. “

What the deal covers

The partnership focuses on next-generation memory design, supply, and manufacturing. NVIDIA said the work will support its next wave of AI systems and expand SK hynix’s role across data centers, personal AI, and physical AI.

The multiyear agreement is designed to support supply for advanced memory, addressing extended development cycles, advanced fabrication, and the heavy capital investments needed for the global AI factory buildout.

NVIDIA and SK hynix will use CUDA-X, PhysicsNeMo, Omniverse, OpenUSD and cuOpt to speed chip development and factory automation. SK hynix is using CUDA-X and AI to speed semiconductor simulation, including technology computer-aided design and computational lithography workflows. The company is also using PhysicsNeMo to accelerate its in-house simulation codes and AI physics workflows.

Additionally, SK hynix is developing fab digital twins to underpin autonomous fab operations. Teams can use Omniverse libraries and OpenUSD pipelines to build 3D factory scenes for visualizing, simulating and optimizing complex semiconductor manufacturing environments. These digital twins can support operational optimization, including the movement of autonomous mobile robots and other fab assets, using cuOpt and the NVIDIA Metropolis platform.

Why the partnership matters

Memory supply has become a constraint in AI hardware. In a recent Reuters article, Huang said the shortage could last for “quite a few years,” which helps explain the push for tighter coordination with suppliers like SK hynix.

The first phase of the AI boom centered on GPUs, but the next phase depends on the full system. Memory, networking, packaging, and manufacturing capacity all shape throughput and cost. As AI models grow larger, memory moves more data and plays a bigger role in performance.

This deal ensures that NVIDIA gains greater influence over the hardware stack supporting its AI systems. SK hynix, meanwhile, gains a deeper role in one of the most important semiconductor markets.

“SK hynix and NVIDIA have been building toward this for years, and this partnership reflects the depth of that collaboration,” said Chey Tae-won, Chairman of SK Group. “Together, we are codeveloping the next generation of memory for AI factories and applying AI to how we design and manufacture semiconductors — work that will shape the future of AI infrastructure.”

Who is SK hynix?

SK hynix is one of the world’s leading memory chip makers and a major supplier of high-bandwidth memory, or HBM. That makes it a critical player in the AI hardware market, where fast memory helps advanced chips move data more efficiently.

The company has spent years building its position in AI memory, and that effort now lines up closely with NVIDIA’s next wave of systems. As demand for AI infrastructure grows, suppliers like SK hynix help determine how quickly those systems can scale.

SK hynix operates in a tightly contested market. Samsung and Micron are major memory suppliers, and both are trying to secure more of the HBM market as demand rises. NVIDIA certified SK hynix alongside Samsung and Micron to supply HBM4 for Vera Rubin, which shows how central memory has become to its roadmap.

That mix of competition and dependence gives SK hynix a strong position. While it’s a supplier, it also participates in design and planning discussions for one of the industry’s most important AI platforms.

What’s happening in the memory market

Demand for memory is rising across the industry as AI spending grows. SK hynix plans to double wafer capacity over the next five years, indicating how strongly the company is preparing for that demand.

Unfortunately, memory capacity doesn’t turn on quickly. New fabs take time, advanced memory is hard to ramp, and supply chains are already stretched by AI demand. A long-term partnership with NVIDIA offers both visibility and planning power.

HBM4 for NVIDIA’s Vera Rubin platform has also become one of the most closely watched supply contests in the industry. Reports have pointed to SK hynix as a leading supplier in that mix, while Samsung and Micron push for share.

The timing of memory supply can shape when AI systems reach the market. According to NVIDIA, this partnership is meant to help align memory supply with the pace of AI infrastructure growth.

What this means for the industry

The deal points to a deeper trend in AI hardware. Major players are coordinating more closely on design, supply, and manufacturing as the industry tries to keep up with demand.

It also raises the stakes for competitors. While Samsung and Micron remain in the HBM market, NVIDIA’s partnership with SK hynix adds another layer of competition to future memory supply.

The AI race is now about speed and supply. It is also about who can supply memory, build fabs, and support the infrastructure that enables chips to be useful at scale.

This deal drew attention because it shows the next stage of AI depends on industrial scale as much as on chip design.

Our take on the deal

The partnership between these companies looks less like a routine supplier update and more like a strategic alignment around the next phase of AI infrastructure. NVIDIA gains greater certainty about a critical input. SK hynix gets a stronger role in one of the most important markets in semiconductors.

The deal also points to a deeper shift in AI hardware. As systems grow more demanding, the companies that control memory and manufacturing will shape the market's pace.