Nvidia’s N1X, the chipmaker’s first serious push into Windows on Arm, leaked in full just days ahead of its expected June 1 reveal at Computex 2026. Specs reported by hardware leak site VideoCardz peg the top chip as a 20-core Arm system on a chip (SoC, the single die that holds CPU, GPU and memory controllers) built from ten Cortex-X925 performance cores and ten Cortex-A725 efficiency cores, wired to a Blackwell graphics engine carrying 6,144 CUDA cores inside a 45-watt to 80-watt power band. Nvidia has not confirmed the numbers.
For about four years, Windows on Arm meant exactly one silicon supplier, Qualcomm. The N1X is the clearest sign yet that the single-vendor era is ending, and Nvidia is walking in with something Qualcomm never put in a thin laptop: a near desktop-class graphics processor bolted to the same package as the CPU.
What the N1X Leak Puts on the Table
The leak does not describe one chip. It describes a four-rung ladder split across two families, with the powerful N1X on top and a leaner N1 below it. Every CUDA core count traces back to the streaming multiprocessor (SM, the basic compute block of an Nvidia GPU), and a Blackwell SM holds 128 CUDA cores. Multiply that out and the leaked figures line up cleanly: 48 SMs gives you 6,144 cores, 40 gives 5,120, and so on down the stack.
| Config | CPU cores (perf + eff) | GPU (SMs / CUDA cores) | Power |
|---|---|---|---|
| N1X (top) | 20 (10 + 10) | 48 / 6,144 | 45-80W |
| N1X (cut-down) | 18 (9 + 9) | 40 / 5,120 | 45-80W |
| N1 (higher) | 12 (8 + 4) | 20 / 2,560 | 18-45W |
| N1 (base) | 10 (7 + 3) | 16 / 2,048 | 18-45W |
The N1 family aims lower on power, sitting in an 18-watt to 45-watt window that overlaps with thin-and-light ultraportables. The N1X reaches into gaming-laptop and small-desktop territory. That spread matters, because it means Nvidia is not chasing a single niche; it is sketching a full product line the way Intel and AMD do.
Why Qualcomm’s Head Start Just Ended
When Microsoft rebuilt Windows for Arm-based chips, it leaned on a single partner. Qualcomm shipped the Snapdragon X line and, for a stretch, had the platform to itself under an exclusivity deal with Microsoft.
That deal ran out. Arm chief executive Rene Haas confirmed the Qualcomm exclusivity agreement was set to lapse, and once it did the door opened for MediaTek, Nvidia and eventually AMD to ship Arm silicon for Windows. The N1X is the loudest knock on that open door. Reports that Nvidia’s first Windows PC chips will debut at Computex, with machines expected from Microsoft’s own Surface line plus Dell and other makers, point to an ecosystem that no longer revolves around one vendor.
Here is what the shift actually changes for buyers and builders:
- Choice arrives where there was none; OEMs can now pit Arm suppliers against each other on price and performance.
- Microsoft gets a second flagship partner pushing app developers to compile native Arm versions.
- Nvidia brings a graphics story to a platform that, until now, sold mostly on battery life and AI inference.
The GPU Nvidia Is Folding Into a Laptop Chip
Strip away the core counts and the headline is the graphics. A discrete laptop GPU in this class is normally a separate chip with its own power budget and cooling. Nvidia’s leaked top N1X folds 6,144 CUDA cores of Blackwell straight onto the processor package, which is the kind of integrated graphics no Arm Windows chip has offered before.
The MediaTek Connection
Nvidia is not doing this alone. At CES 2025 the company detailed a collaboration with MediaTek on Arm-based PC silicon, building on the GB10 Grace Blackwell Superchip that powers its Project DIGITS desktop box. Nvidia’s Grace Blackwell developer platform announcement framed the pairing as Nvidia’s accelerated computing on top of MediaTek’s CPU power efficiency, and MediaTek’s GB10 Superchip briefing confirmed it co-designed that chip’s Arm CPU.
The combination of MediaTek’s industry-leading CPU performance and power efficiency with NVIDIA’s accelerated computing technologies will drive the next wave of innovation.
That line came from Jensen Huang, founder and chief executive of Nvidia, describing the partnership at CES. The N1X reads like the consumer-laptop expression of that same Grace-plus-Blackwell formula.
What the Core Count Buys
Big integrated graphics unlocks things Snapdragon cannot match today: hardware ray tracing, the full RTX feature set, and DLSS (Deep Learning Super Sampling, Nvidia’s AI frame-upscaling tech) inside a fanless or single-fan chassis. For gaming and on-device AI work, that is a genuine capability gap, not a spec-sheet flex. The catch is that all of it still runs on top of Windows on Arm, where the software has to cooperate.
Where Snapdragon, Core Ultra and Ryzen Sit
Nvidia is not arriving to an empty room. Qualcomm’s Snapdragon X2 Elite Extreme already runs an 18-core Arm design clocking up to 5.0 GHz with an 80 TOPS NPU (neural processing unit, the block that runs AI tasks; TOPS means trillion operations per second). On the x86 side, Intel still holds roughly 74% of laptop CPUs, and AMD pushed to a record 29.2% share of the broader x86 market in late 2025, per worldwide x86 laptop share data.
The pressure on each incumbent looks different:
- Qualcomm loses its status as the only Arm game in Windows and now has to defend on graphics, where it is weakest.
- Intel faces an Arm rival aimed straight at the thin-and-light volume that anchors its laptop lead.
- AMD sees its Ryzen integrated-graphics edge, long its differentiator, challenged on a platform it has not yet shipped into.
The runway is real. Arm-based PCs are projected to take a meaningful slice of the market through this year, with Arm Windows laptops expected to reach 8 to 10% of the Windows PC market by the end of 2026. A credible Nvidia entrant pointed at that growth is exactly what the x86 camp did not want to see.
The 2027 Catch in a 2026 Reveal
Here is the part the spec sheet does not shout. A Computex reveal is not a Computex launch. Industry tracking firm TrendForce reported that the N1X timeline has slipped more than once, and current expectations put commercial machines in the first quarter of 2027, not on store shelves this summer.
The reasons stack up:
- Chip revisions, including reported fixes to underlying logic design and the chip’s metal-layer interconnects, that forced the silicon back into a respin.
- Slow progress on Microsoft’s next-generation Windows on Arm software, which was meant to land alongside the chip.
- Lingering bugs and softer notebook demand that gave both sides reason to wait.
There is also the app problem that has dogged the platform from day one. Windows on Arm runs x86 software through emulation, and a handful of pro apps, drivers and anti-cheat systems still stumble. A monster integrated GPU does not fix a game that refuses to launch. Nvidia gets to dazzle the Computex floor the way every vendor does at the show, but the gap between a demo and a shipping laptop is where this story actually gets decided.
If the silicon Nvidia shows on June 1 ships on schedule in early 2027, Qualcomm spends roughly a year as the only Arm option before the real fight starts. If the chip slips again, that head start stretches into a second year Qualcomm never had to earn.
Frequently Asked Questions
What are the full Nvidia N1X specs?
The leaked top N1X carries a 20-core Arm CPU (ten Cortex-X925 performance cores plus ten Cortex-A725 efficiency cores) and a Blackwell GPU with 48 SMs, equal to 6,144 CUDA cores, in a 45-watt to 80-watt envelope. A cut-down version drops to 18 cores and 5,120 CUDA cores. Nvidia has not officially confirmed these figures.
When will the Nvidia N1X launch?
Nvidia is expected to reveal the chip around June 1 at Computex 2026, but commercial laptops are not projected until the first quarter of 2027 after repeated delays tied to chip revisions and Windows on Arm software readiness.
How is the N1 different from the N1X?
The N1 is the cheaper, lower-power family aimed at thin-and-light laptops, with 12- and 10-core CPU options and 2,560 or 2,048 CUDA cores in an 18-watt to 45-watt range. The N1X is the high-performance line with far more cores, more graphics and a higher power ceiling.
Can the N1X run normal x86 Windows apps?
Yes, through emulation. Windows on Arm translates x86 software to run on Arm chips, but some professional apps, drivers and anti-cheat systems still have compatibility gaps, which remains the platform’s biggest limitation regardless of how powerful the hardware is.
Did Nvidia build the N1X with MediaTek?
Yes. Nvidia confirmed a collaboration with MediaTek on Arm-based PC silicon at CES 2025, pairing Nvidia’s GPU and accelerated computing technology with MediaTek’s CPU power-efficiency work, the same approach behind the GB10 Grace Blackwell Superchip.
Does the N1X threaten Qualcomm’s Snapdragon X chips?
It targets them directly. Qualcomm lost its exclusive hold on Windows on Arm when its Microsoft deal expired, and Nvidia’s integrated Blackwell graphics give the N1X a clear advantage in gaming and on-device AI that current Snapdragon X parts cannot match.
