NVIDIA CEO Jensen Huang is set to unveil next-generation artificial intelligence (AI) PC chips in Taiwan, featuring SK Hynix's LPDDR5X memory. Following its success in high-bandwidth memory (HBM) for servers, SK Hynix is accelerating its integration into NVIDIA's ecosystem in the AI PC market.
According to industry sources, Huang will focus on the AI chips 'N1' and 'N1X' during his keynote address at GTC 2026 in Taiwan. These products are expected to be system-on-chip (SoC) solutions that combine Arm-based central processing units (CPUs) with NVIDIA graphics processing units (GPUs). Reports indicate that SK Hynix's LPDDR5X will serve as the primary memory supply for these chips.
The chips are anticipated to support up to 128GB of LPDDR5X memory, significantly larger than typical smartphones or existing laptops. To run generative AI and on-device AI functions directly on laptops, both CPU and GPU performance, as well as memory capacity and bandwidth for processing large datasets, are crucial. This demand for AI PCs is directly linked to the growing need for high-performance, low-power DRAM.
The N1 and N1X are regarded as NVIDIA's first serious entries into the AI PC market. The N1 targets the premium AI laptop segment, while the N1X is aimed at high-performance laptops and mobile workstations. Specific details regarding core counts, power design, and GPU configurations are expected to be revealed in an official announcement.
A key feature of these chips is the integration of CPU and GPU within a single package. Traditionally, the PC market has been dominated by Intel and AMD, with NVIDIA enhancing performance through discrete GPUs. In contrast, the N1 and N1X aim to redefine the fundamental computing architecture of PCs by combining Arm CPUs with NVIDIA GPUs, focusing on AI.
The competitive landscape for AI PCs may also shift. Previously, laptop comparisons centered on CPU performance, battery efficiency, and graphics capabilities. However, with generative AI becoming integral to PCs, the ability to execute large language models, generate images and videos, assist with coding, and analyze documents will become key competitive factors. The N1 and N1X are positioned as products that unify CPU, GPU, and memory into a single AI computing platform.
Some in the semiconductor industry speculate that these products could disrupt the Intel-centric PC ecosystem. Arm-based chips have already proven their power efficiency in the smartphone and tablet markets. The combination of NVIDIA GPUs and AI software could create a new competitive axis in the Windows PC market. Analysts view NVIDIA's collaboration with the Arm ecosystem as a significant challenge to the x86-dominated PC market.
For SK Hynix, the introduction of LPDDR5X alongside HBM offers new opportunities. The company has improved its performance in the AI server market through HBM supply, and expanding LPDDR5X supply for AI PCs would position it within NVIDIA's value chain on both server and client sides.
The integration of 128GB LPDDR5X is particularly significant for memory manufacturers. This capacity far exceeds the typical LPDDR amounts found in smartphones. As AI PC shipments increase, the rise in memory capacity per unit could lead to substantial revenue growth. Unlike traditional PC DRAM demand, which has stagnated due to slower replacement cycles, AI PCs require high-capacity memory, presenting a new market opportunity.
The most direct beneficiary of the N1 and N1X project is likely to be SK Hynix, as NVIDIA has elevated HBM to a near-essential component in the AI server market. Similarly, the importance of high-capacity, low-power memory in AI PCs is expected to grow. As AI processing shifts from the cloud to PC devices, technologies that reduce memory bottlenecks will become critical to product performance.
Recently, SK Hynix memory was confirmed on NVIDIA's N1 and N1X sample boards, increasing the likelihood of supply.
In response, a representative from SK Hynix stated, "We cannot confirm details regarding collaborations with our clients due to security regulations."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.