Nvidia has spent years teasing its way into the laptop market, and now — one day before Jensen Huang takes the Computex stage in Taipei — a Geekbench result filed from a mystery HP prototype has given the internet its first hard look at the N1X SoC's CPU performance. The scores are good. They're also running Linux on pre-production silicon with 128 GB of RAM. That combination of factors matters enormously for how you interpret them.
The listing, filed under 'NVIDIA N1x' on Geekbench 6.4.0 Preview for Linux AArch64, identifies a 20-core ARM v8 chip at a 2.81 GHz base frequency, paired with what the tool reports as 119.59 GB of memory — almost certainly a 128 GB LPDDR5X pool being rounded imprecisely.
The motherboard is logged as HP 8EA3, an internal prototype designation that confirms this is engineering hardware, not a shipping product. The result: 2,821 single-core and 17,152 multi-core.
Geekbench 6 — CPU Comparison
| Chip | Single-Core | Multi-Core |
|---|---|---|
| Nvidia N1X (Linux, pre-prod) | 2,821 | 17,152 |
| AMD Ryzen AI MAX+ 395 | 3,125 | 21,035 |
| Intel Core Ultra 9 285HX | 3,078 | 22,104 |
| Qualcomm Snapdragon X Elite | 2,693 | 13,950 |
| Apple M4 Max (macOS) | 4,054 | 25,913 |
What the chip actually is
The N1X is Nvidia's first SoC designed specifically for Windows ARM laptops, and it is architecturally unlike anything currently available. Built on TSMC's 3nm process, it marries a CPU die co-developed with MediaTek — 10 Cortex-X925 performance cores plus 10 Cortex-A725 efficiency cores — to a Blackwell GPU die carrying 6,144 CUDA cores across 48 streaming multiprocessors.
That GPU core count matches Nvidia's desktop RTX 5070. The two dies connect via Nvidia's NVLink C2C interconnect at 300 GB/s bidirectional bandwidth, and the chip supports up to 128 GB of shared LPDDR5X memory.
The benchmark caveat no one is discussing loudly enough
The Geekbench result surfaces from a Linux environment. That is a meaningful asterisk. Linux typically yields higher scores on ARM silicon than Windows 11 does, largely because Windows on ARM incurs overhead from the x86 emulation layer (Prism) and because mature driver stacks have not yet shipped.
When Qualcomm's Snapdragon X Elite was benchmarked on Linux before its Windows laptop launch, it scored notably higher than it achieved in production hardware. The N1X multi-core result of 17,152 sits roughly 10–15 percent behind AMD's Ryzen AI MAX+ 395 (21,035) and Intel's Core Ultra 9 285HX (22,104) — but those comparisons ran on Windows 11, not Linux, which further distorts the gap.
"These are prototype scores from pre-production silicon on Linux. Real Windows laptop results will almost certainly differ, and finalized drivers could move performance in either direction." — Context note — scores should be treated as directional, not final
Where Nvidia's real advantage lives
CPU figures are arguably the least interesting part of the N1X story. The integrated Blackwell GPU is estimated to deliver laptop performance comparable to the RTX 4070 and RTX 5070 in compute workloads — significantly ahead of any existing integrated graphics in the Windows ARM space.
More critically, it brings full CUDA support to a thin-and-light laptop for the first time. That ends the long-standing dilemma for AI developers who had to choose between CUDA workflows (on high-performance gaming laptops or in the cloud) and Apple Silicon's efficiency. The N1X potentially supports locally run inference for 13B to 70B-parameter models on a device that fits in a bag.
"From an industry perspective, it's a good thing. Qualcomm has struggled to grab a significant chunk of the PC market despite offering excellent battery life, in part because developers didn't see a need to focus scarce resources on a somewhat different version of Windows." — Carolina Milanesi, analyst at Current Strategies, speaking to Axios
Which laptops are coming
Dell's XPS line, Lenovo's Legion 7, ASUS ProArt, and Microsoft's Surface brand have all confirmed or strongly signaled N1X and N1 variants ahead of Computex. Jensen Huang's GTC Taipei keynote is set for June 1 at 11am local time.
Microsoft will pair the hardware launch with new software that enables AI agents to run locally on Windows — a direct response to spiraling cloud compute costs that enterprises are increasingly pushing back against.
Whether the N1X reshapes the Windows laptop market or settles into a similar niche position to Qualcomm's Snapdragon X series comes down almost entirely to developer adoption. NVIDIA's name recognition and the CUDA ecosystem give it a pull that Qualcomm never had. That is the variable Intel and AMD should be watching most carefully when those keynote slides hit the screen tomorrow.