NVIDIA’s highly anticipated N1X System-on-Chip (SoC), central to its Windows on Arm (WoA) strategy, is now reportedly slated for a Q1 2026 launch, a delay from its original 2H 2025 target. This pushback is attributed to a confluence of factors, including Microsoft’s decelerating development roadmap for its next-generation Windows platform, a slower-than-anticipated adoption rate for WoA devices, and NVIDIA’s internal evaluation of new revisions for its N1X and N1 SoCs.
The N1X SoC, co-developed with Mediatek, marks NVIDIA’s significant entry into the enterprise and consumer “AI PC” segment. While NVIDIA and its partners, including ASUS, MSI, Gigabyte, Lenovo, and Dell, showcased various devices, such as NVIDIA’s own DGX Spark, at Computex 2025, the actual market availability has been postponed. The delay in Microsoft’s Windows OS update, which is expected to introduce advanced AI features and enhance the WoA ecosystem, is a primary reason for NVIDIA’s adjusted timeline. This critical software foundation is deemed essential for the optimal performance and market reception of the N1X.
Furthermore, the broader market adoption of WoA platforms has not matched initial expectations. Despite claims of strong sales from some manufacturers, overall consumer demand for WoA laptops is reportedly slowing. The “AI PC” narrative, while continuously pushed, has yet to fully translate into widespread consumer purchases, with x86 platforms from AMD and Intel continuing to dominate the market share.
Internally, NVIDIA is leveraging this extended timeline to refine its N1X and N1 SoCs. The company is evaluating new revisions, which are expected to incorporate design changes and various optimizations. These chips are being manufactured using TSMC’s N3B process technology. The N1X SoC integrates a Grace (Arm) CPU and a Blackwell GPU, also referred to as GB10 in platforms like the DGX Spark. The N1X designation is specifically tailored for the consumer market.
In terms of specifications, the NVIDIA GB10 Superchip, which shares many characteristics with the N1X SoC, features a 20-core Arm CPU based on the Grace architecture. It also includes a Blackwell GPU capable of up to 1 PetaFlop of FP4 AI compute, supports up to 128 GB of LPDDR5X memory, and incorporates advanced I/O technologies.
Early performance indicators for the N1X SoC are promising. A Geekbench 6 listing for the chip shows a single-core score of 3096 and a multi-core score of 18,837. The tested chip operated at 2.81 GHz, with samples of the GB10 known to reach clock speeds of up to 3.9 GHz. For context, AMD’s Ryzen AI MAX+ 395 chip, with 16 cores and 32 threads, and clock speeds exceeding 5 GHz at over 120W, achieves single-core scores around 3000-3100 and multi-core scores between 19000-21000. This indicates that even in its current early state, the N1X demonstrates competitive performance, with further improvements anticipated from the planned revisions.
The NVIDIA N1 series is expected to encompass multiple chip variants. The N1X is targeted at the Enterprise/Commercial segment, while the N1C will cater to the consumer market. An entry-level model is also anticipated for niche applications, such as handheld devices. While initial expectations for these chips included AI TOPs performance in the range of 180-200, the final configurations and performance metrics are subject to change due to the ongoing design revisions.
Upon its eventual launch, the NVIDIA N1 series will enter a competitive landscape, vying for market share against established players such as AMD’s “Ryzen AI MAX,” Apple’s M4, and Intel’s AX series. These chips are poised to play a significant role in the evolving “AI PC” market in the coming years. Further details regarding the N1X and the broader N1 series are expected to be unveiled at CES 2026.
