Nvidia N1 Chip Leak Points To Major Windows-On-Arm Push
Nvidia’s long-rumoured move into Arm-based PC processors appears to be edging closer, with leaked specifications revealing several variants of its upcoming N1-series chips ahead of an expected launch.
According to documents reported by VideoCardz, Nvidia is preparing at least four N1-series processors, split between high-performance N1X models and lower-power N1 chips aimed at thinner laptops.
The flagship N1X is said to share similarities with Nvidia’s GB10 processor, used in the company’s DGX Spark AI system.
The top model reportedly features a 20-core Arm CPU made up of 10 Cortex-X925 performance cores and 10 Cortex-A725 efficiency cores, paired with Blackwell graphics offering 48 Streaming Multiprocessors, or 6,144 CUDA cores.
A second N1X variant is tipped to carry an 18-core CPU and 5,120 CUDA cores. Both versions are expected to operate in a 45W to 80W power range, putting them in the same territory as premium laptop processors.

The N1X platform could also support up to 128GB of LPDDR5X memory and as many as three M.2 SSDs, suggesting Nvidia is targeting high-end laptops, mobile workstations and AI-focused PCs.
However, the standard N1 chips may prove more important for mainstream PC makers. Leaked configurations point to 12-core and 10-core versions, with 2,560 and 2,048 CUDA cores respectively. These chips are said to run within an 18W to 45W power range and support up to 64GB of LPDDR5X memory.
That positions the N1 range directly against the current wave of Windows-on-Arm laptops, a market largely driven by Qualcomm’s Snapdragon X platform.
The leak also suggests Nvidia has been working on the processors for some time, with at least one slide reportedly dated 2024.
Not all leaked variants are guaranteed to ship and Nvidia has not confirmed the specifications.
But if accurate, the N1 family would mark Nvidia’s most serious attempt yet to enter the PC processor market, combining Arm CPU cores with its own Blackwell GPU technology in a single laptop platform.























































































