Nvidia's elusive N1X SoC leaks out again — shipping manifest reveals Dell may have explored putting it on next-gen XPS laptops

Nvidia GB10
(Image credit: Nvidia)

It's been more than a year since Nvidia's N1/N1X first leaked, yet the rumored chips were still a no-show at CES this year, as expected. Beyond a few rumors, the tight-lipped news cycle surrounding these chips points toward scaled-back ambitions. The latest leak, in particular, might paint a grimmer picture than a hopeful one: a new shipping manifest showing a Dell laptop that's since been entirely rebranded, perhaps hinting at the fate of the chip it was seen with, as well.

The device we're talking about is the Dell 16 Premium, which the brand thankfully renamed back to Dell XPS at CES 2026. It was spotted with an "N1X ES2" in a shipping manifest, meaning that Dell may have been experimenting with engineering samples of the N1X as late as November 20, 2025. Whether those were considerations for final units, or even just a classic CES concept appearance, we don't know.

Leaked shipping manifest showing Nvidia N1X inside a Dell Premium 16 laptop

(Image credit: @Olrak29_ on X)

The fact that the laptop is still mentioned as a Dell 16 Premium this late in its development cycle, instead of an XPS, could indicate a last-minute change. Pessimistically thinking, maybe at one point the N1X was supposed to make it through to retail, but the rebrand killed those aspirations entirely. If this leak is to be believed, both Nvidia and Dell's internal leadership could be responsible for that decision. Right now, though, all we have is speculation.

For those out of the loop, for years we've heard about the N1 silicon: purported to be an RTX 5070-class GPU featuring 6,144 CUDA cores, combined with 20 ARM cores built on the Grace architecture. Co-developed with MediaTek, the N1 is actually what powers the GB10 Superchip inside the DGX Spark, as confirmed by Jensen Huang himself. So, the hardware definitely does exist, but it's only served as a blueprint for Nvidia's AI efforts instead of getting a mainstream desktop (N1) or mobile (N1X) release.

If launched, the N1 lineup would mark Nvidia's grand return to the conventional CPU segment in more than a decade, since the Tegra X1 last powered the Nintendo Switch. To be clear, the company never stopped making ARM-based CPU but they've been limited to enterprise-grade solutions, not something consumer-facing like, say, a Ryzen APU.

That's the market the N1 was purportedly targeting, which is what made the chips so exciting. Conversely, Windows-on-ARM isn't the most exciting thing ever, and the N1's graphical prowess might've even been lost on it, but it would've certainly progressed the platform. Rivals like Qualcomm are already working to make ARM-based systems run Windows without a hiccup.

There's also a small, conflict-of-interest-sized elephant in the room, which is Nvidia's equity investment in Intel. The two are in partnership to co-develop Intel x86 RTX SoCs, likely using high-bandwidth interconnects to put RTX chiplets next to Intel CPUs. So far, reports have said these efforts coincide with each other, but if an existing product never transitions from blueprint to retail, it wouldn't be difficult to guess why.

Google Preferred Source

Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

Hassam Nasir
Contributing Writer

Hassam Nasir is a die-hard hardware enthusiast with years of experience as a tech editor and writer, focusing on detailed CPU comparisons and general hardware news. When he’s not working, you’ll find him bending tubes for his ever-evolving custom water-loop gaming rig or benchmarking the latest CPUs and GPUs just for fun.

  • bit_user
    The article said:
    For those out of the loop, for years we've heard about the N1 silicon: purported to be an RTX 5070-class GPU featuring 6,144 CUDA cores, combined with 20 ARM cores built on the Grace architecture. Co-developed with MediaTek, the N1 is actually what powers the GB10 Superchip inside the DGX Spark, as confirmed by Jensen Huang himself. So, the hardware definitely does exist, but it's only served as a blueprint for Nvidia's AI efforts instead of getting a mainstream desktop (N1) or mobile (N1X) release.
    So, I've tried to follow the rumors and it's not clear to me whether the GPU die in N1(X) is the same as GB10, or just the CPU portion.

    As explained here, the GB10 is actually comprised of two logic dies: the CPU cores (and SoC) + GPU die:

    Source: https://www.chiplog.io/p/analysis-of-nvidia-dgx-sparks-gb10
    Note that the block diagram isn't to scale. In the GB10, the two dies are closer to equal in size.

    What we know is that the N1(X) features the same 20-core configuration and we've heard that it's the same silicon, but I don't know if any specs have leaked about the N1(X) GPU portion. Anyway, if you think about it, it would make a lot of sense for them to differentiate by swapping in different GPU tiles.
    Reply
  • bit_user
    The article said:
    If launched, the N1 lineup would mark Nvidia's grand return to the conventional CPU segment in more than a decade, since the Tegra X1 last powered the Nintendo Switch.
    Uh, what are you talking about??
    Have you heard of the Switch 2? You know Nvidia makes that SoC, as well, right? They have an even newer SoC than the one in Switch 2! These are primarily aimed at embedded applications, like robotics and self-driving cars, but they're fully capable of being used in client devices.
    https://en.wikipedia.org/wiki/Tegra#Orin
    What's notable about the GB10 and N1(X) is that it's actually a partnership with MediaTek, rather than Nvidia running the whole show on its own. It's a little hard to see what Nvidia gets out of sharing the glory with MediaTek, but then I guess it does save Nvidia the burden of having to support the SoC on Windows.

    Also, as noted in the article, expect to see a similar integration model between Nvidia iGPUs and future Intel laptop CPUs.
    Reply