Skip to content
AI Tech

NVIDIA Just Built a Laptop Chip. Here’s Why That’s a Big Deal

✅ Updated — June 1, 2026

Jensen Huang’s Computex 2026 keynote has taken place. The N1X is now officially announced under the consumer brand name RTX Spark. Confirmed OEM partners expanded to include HP alongside Dell, Microsoft, ASUS, Lenovo and MSI. All confirmed specs below have been updated.

✅ Officially Announced · Computex 2026 · June 1

NVIDIA N1X / RTX Spark — First Laptop Chip: Everything Officially Confirmed

NVIDIA officially unveiled its first laptop chip at Computex 2026 on June 1, 2026 — branding it the RTX Spark. The chip pairs a 20-core Arm CPU with Blackwell GPU and 6,144 CUDA cores, equivalent to RTX 5070-class desktop graphics. Confirmed OEM partners: Microsoft, Dell, HP, ASUS, Lenovo and MSI. First devices before holiday 2026.

By Mr Wangdoo · 30 May 2026 · Updated 1 June 2026 — post-keynote confirmed

Editorial transparency: This article was written on 30 May 2026 ahead of Jensen Huang’s Computex 2026 keynote and updated on June 1, 2026 following the official announcement. Every claim is sourced from confirmed OEM leaks, VideoCardz reporting, TechTimes analysis, Tom’s Hardware, TweakTown, Resetera hardware forums, and official teasers from NVIDIA, Microsoft, Arm, and MediaTek. Specs marked CONFIRMED come from multiple independent sources. Specs marked RUMOURED have not been officially announced by NVIDIA. Confirmed specs are now marked accordingly. Wangdoo.com has no commercial relationship with NVIDIA, Microsoft, MediaTek, or Arm.

On May 29, 2026, four companies posted an identical message simultaneously. NVIDIA, Microsoft’s Windows account, Arm, and MediaTek all published the exact same text — “A new era of PC.” — followed by the GPS coordinates 25.0528, 121.5990. Those coordinates map to the Taipei Music Center, where NVIDIA CEO Jensen Huang delivered his GTC Taipei keynote on June 1, 2026. CNBC confirmed from the keynote that the chip is officially branded the RTX Spark — NVIDIA’s consumer name for the N1X SoC — and that it will debut on laptops from Microsoft, Dell, HP, ASUS, Lenovo and MSI.

The coordinated post was not a coincidence. It was a choreographed signal from the four companies that together represent the entire supply chain of what is about to be announced: NVIDIA’s N1 and N1X — the company’s first laptop processor chips, and its first serious move into the consumer CPU market in over two decades.

What This Article Covers

Everything officially confirmed from Jensen Huang’s Computex 2026 keynote — including the RTX Spark brand name, confirmed OEM laptop models, specific device names, Jensen Huang’s own quotes, and how RTX Spark compares to Qualcomm’s Snapdragon X and Apple Silicon.

20-coreArm CPU (N1X)
6,144CUDA cores — Blackwell GPU
RTX 5070Estimated GPU class
4+Confirmed OEM partners
Jun 1Official announcement keynote
TSMC 3nmManufacturing process

Jensen Huang speaks in Taiwan ahead of Computex 2026 — update: full keynote available June 1

What Are the NVIDIA N1 and N1X?

The N1 and N1X are NVIDIA’s first system-on-chip (SoC) processors designed specifically to go inside Windows laptops. This is significant because NVIDIA has spent the past three decades as a GPU company — making graphics cards that sit alongside someone else’s CPU. The N1 series changes that entirely. For the first time, an NVIDIA chip will handle both processing and graphics inside a consumer laptop.

The chips are co-developed with MediaTek — Taiwan’s largest chipmaker and a major supplier to the mobile industry. MediaTek contributed the Arm CPU design and manufacturing coordination, while NVIDIA brings its Blackwell GPU architecture and the full CUDA software stack. The combined result is a chip that targets the same market as Apple’s M-series Silicon and Qualcomm’s Snapdragon X — powerful, efficient, Arm-based computing — but with NVIDIA’s graphics and AI credentials behind it.

✓ Confirmed by multiple independent sources

The N1X chip name appears in NVIDIA’s driver stack (spotted in the 560.x branch by driver researchers) and in Windows 11 Insider builds, confirming real silicon is in active validation. Lenovo’s internal ADFS authentication system references an “NVIDIA N1x Portal PROD” and “NVIDIA N1x Portal Test” — confirming production hardware is in Lenovo’s internal labs.

RTX Spark (N1X) — Officially Confirmed Specs

Important before reading

All specs below are sourced from Jensen Huang’s official Computex 2026 keynote (June 1), CNBC, Tom’s Hardware, HotHardware, and Notebookcheck post-keynote reporting. Previously rumoured specs that have now been officially confirmed are marked accordingly.

SpecificationN1X DetailStatus
Chip nameNVIDIA N1X / RTX Spark (official consumer brand announced June 1) — N1 is the mainstream variantConfirmed — keynote
CPU cores20-core Arm CPU — 10 performance + 10 efficiencyConfirmed
GPU architectureBlackwell — same architecture as RTX 50-series desktop GPUsConfirmed
CUDA cores6,144 CUDA coresConfirmed via leaks
GPU performance classRTX 5070-class mobile graphics — confirmed equivalent to desktop RTX 5070 by CNBC post-keynoteConfirmed — keynote
NPUDedicated neural processing unit for on-device AI workloadsConfirmed
ManufacturingTSMC 3nm process node — exact variant (N3P or N3E) unconfirmedRumoured
Memory128GB unified memory (LPDDR5X-8533) — confirmed by CNBC and Tom’s Hardware post-keynoteConfirmed — keynote
TDP range65W – 120W depending on configurationRumoured
Co-developerMediaTek (CPU design + manufacturing coordination)Confirmed
CUDA software compatibilityFull CUDA stack — day-one developer tool compatibility expectedConfirmed via teaser
Windows compatibilityWindows on Arm — Windows 11 nativeConfirmed
Power target (N1X)High-performance — Legion 7 charger listing shows 245W adapterRumoured
PricingNot confirmed — LPDDR5X and TSMC 3nm suggest premium positioningUnconfirmed
AvailabilityFall 2026 — more than 30 laptops and 10 desktops confirmed by NVIDIAConfirmed — keynote

Confirmed OEM Partners and Devices

Before a single official word from NVIDIA, at least four major OEMs have confirmed or strongly indicated N1/N1X products through Computex media materials, internal leaks, and public teasers:

Dell XPS laptop with N1X — embargoed product set for May 31 reveal. Confirmed via Computex media directory. Confirmed
Lenovo Legion 7 15N1X11 (gaming, 245W charger), IdeaPad Slim 5 (14″ + 16″), Yoga Pro 7 (15″), Yoga 9 2-in-1 (16″) — leaked via internal ADFS system and support pages. Confirmed via leak
ASUS ProArt laptop targeting creative professionals — teased publicly by ASUS on X with a winking emoji post on May 29. Confirmed
Microsoft Surface variant with N1X / RTX Spark — confirmed by Jensen Huang’s June 1 keynote. Microsoft participated in the coordinated “new era of PC” teaser on May 29. Confirmed — keynote
HP HP laptop with N1X / RTX Spark — confirmed by CNBC reporting from the June 1 keynote. HP was not in pre-announcement leaks but was included in the official OEM partner announcement. Confirmed — keynote
MSI At least one N1X system in preparation — confirmed by VideoCardz sources ahead of Computex. Rumoured

Jensen Huang’s Own Words — Official Quotes From the Keynote

“This reinvention of the computer is as big of a deal as the reinvention of the phone into what we now know as the smartphone.”

— Jensen Huang, CEO NVIDIA · Computex 2026 keynote, June 1, 2026

“Microsoft and NVIDIA are going to reinvent the PC. This is the first completely re-engineered, reinvented line of PCs that has happened in 40 years.”

— Jensen Huang, CEO NVIDIA · Computex 2026 keynote, June 1, 2026

NVIDIA’s initial plan is to release more than 30 laptops and 10 desktops with the RTX Spark chip. Huang also confirmed a joint keynote with Microsoft CEO Satya Nadella on June 2, 2026 — where further details about the Windows and NVIDIA partnership will be revealed.

Confirmed Device Models at Launch

Notebookcheck and Computex press materials confirm the following specific models launching with RTX Spark:

  • ASUS ProArt P14 — creative professional laptop
  • ASUS ProArt P15 — creative professional laptop
  • Dell XPS 16 — premium consumer laptop
  • HP OmniBook X 14 — mainstream laptop
  • HP OmniBook Ultra 16 — premium laptop
  • Lenovo Yoga Pro 9n — premium 2-in-1
  • Microsoft Surface Laptop Ultra — premium Surface
  • MSI Prestige N16 Flip AI — creator/business convertible

NVIDIA also confirmed RTX Spark will power some mini-PCs, though specific models were not named at the keynote. More than 30 total laptop models are planned across all OEM partners.

To understand why the N1X is significant, it helps to understand what the laptop chip market has looked like for the past three years. Apple’s M-series Silicon — starting with the M1 in 2020 — demonstrated that Arm-based processors could outperform x86 Intel chips while using a fraction of the power. The laptop industry has been scrambling to respond ever since.

Qualcomm’s Snapdragon X, launched in 2024, was the first serious Windows-on-Arm alternative. It delivered strong CPU performance and battery life, but its integrated GPU lagged behind what gamers and creative professionals needed. The gap between “good for productivity” and “good for serious GPU workloads” remained wide.

The N1X closes that gap in a way no previous Windows-on-Arm chip has. By combining MediaTek’s Arm CPU with NVIDIA’s Blackwell GPU architecture and the full CUDA software stack, it targets a laptop that can handle both all-day productivity and serious local AI workloads — including running large language models, generating images, and training small models entirely on-device without cloud access.

The CUDA Advantage

CUDA is NVIDIA’s proprietary computing platform — the software stack that most AI research tools, deep learning frameworks, and developer tools are built on. PyTorch, TensorFlow, Stable Diffusion, Ollama, and thousands of other AI tools are CUDA-native. Bringing CUDA to a thin-and-light Windows laptop means AI developers who previously needed a cloud instance or a gaming laptop to run serious workloads could do it on a portable machine for the first time.

N1X vs Qualcomm Snapdragon X vs Apple M4 — How They Compare

FeatureNVIDIA N1XQualcomm Snapdragon X EliteApple M4 Pro
CPU architectureArm (20-core)Arm (12-core Oryon)Arm (14-core)
GPU architectureBlackwell (NVIDIA)Adreno X1-85Apple GPU (20-core)
CUDA supportYes — full stackNoNo — Metal only
RTX / ray tracingYes (Blackwell)NoNo
Local AI (LLM) performanceRTX 5070-class — confirmed (no dGPU; strong local AI focus)Capable but GPU-limitedStrong via unified memory
Windows compatibilityWindows 11 nativeWindows 11 nativemacOS only
ManufacturingTSMC 3nmTSMC 4nmTSMC 3nm
Expected battery lifeNot yet confirmed20+ hours (Copilot+ PCs)18–24 hours
Gaming performanceRTX 5070-class integrated — no discrete GPU support confirmedModerateModerate
AvailabilityLate 2026 / Early 2027Available nowAvailable now
Price rangePremium — unconfirmed$999–$1,799From $1,599 (M4) / $1,999 (M4 Pro)

Who Is the N1X Built For?

Based on confirmed specs and the OEM lineup, the N1X targets four distinct audiences:

  • AI developers and researchers — who currently need a cloud GPU or a heavy gaming laptop to run CUDA-native AI workflows. The N1X brings the full CUDA stack to a portable Windows machine for the first time.
  • Creative professionals — ASUS ProArt targeting this market suggests NVIDIA and partners see video editing, 3D rendering, and image generation as primary workloads. RTX 5070-class GPU performance is transformative for creative work on a laptop.
  • Gamers — the Lenovo Legion 7 N1X with a 245W charger signals a gaming-class configuration. RTX Blackwell with full ray tracing and DLSS 5 on an Arm Windows laptop is an entirely new product category.
  • Privacy and local AI users — anyone running tools like Ollama, Stable Diffusion, or local LLMs who wants to move off cloud infrastructure. N1X + CUDA = the fastest local AI on a portable Windows device.

Frequently Asked Questions

What is the NVIDIA N1X?
The NVIDIA N1X is NVIDIA’s first laptop system-on-chip (SoC), co-developed with MediaTek. It combines a 20-core Arm CPU with NVIDIA’s Blackwell GPU architecture and 6,144 CUDA cores — bringing RTX-class graphics and the full CUDA software stack to Windows-on-Arm laptops for the first time. It is being officially announced by CEO Jensen Huang at GTC Taipei on June 1, 2026, one day before Computex 2026 opens.
What does “A new era of PC” mean — who posted it?
On May 29, 2026, NVIDIA, Microsoft’s Windows account, Arm, and MediaTek simultaneously posted identical messages stating “A new era of PC” alongside the GPS coordinates 25.0528, 121.5990 — which map to the Taipei Music Center, the venue for Jensen Huang’s Computex 2026 keynote. The coordinated post from all four companies — chip designer, OS maker, CPU architecture licensor, and manufacturing partner — was a deliberate signal that a major joint announcement was coming. That announcement is the N1 and N1X chips.
How does the NVIDIA N1X compare to Apple M4?
The N1X and Apple M4 are both Arm-based, TSMC-manufactured SoCs targeting high-performance portable computing. The key differences: the N1X runs Windows and supports CUDA — making it compatible with the vast ecosystem of CUDA-native AI, gaming, and creative tools. The Apple M4 runs macOS and uses Apple’s Metal framework, which excludes it from the majority of CUDA-dependent AI research tools. On GPU performance, the RTX Spark is confirmed at RTX 5070 class — significantly ahead of the M4 Pro’s integrated GPU. Apple’s advantage remains battery life and macOS optimisation, both of which are proven. N1X battery life remains unconfirmed.
Will the NVIDIA N1X run games?
Yes — this is one of the most significant aspects of the N1X. The Blackwell GPU with 6,144 CUDA cores and confirmed RTX 5070-class performance brings full ray tracing, DLSS 4.5 Ray Reconstruction, and RTX features to a Windows-on-Arm laptop. Note: RTX Spark has no discrete GPU capability — NVIDIA’s focus is agentic AI and mainstream gaming, not high-end gaming workloads. The Lenovo Legion 7 N1X configuration — with a 245W charger — signals a gaming-class thermal and power envelope. Windows gaming on Arm has improved dramatically over the past two years, making this a viable gaming platform for the first time.
When will NVIDIA N1X laptops be available to buy?
Based on OEM timelines and supply chain reports, the first N1X laptops are expected to reach market before the 2026 holiday season — broadly meaning October to December 2026. Broader availability, including more affordable models and wider retail presence, is expected in early 2027. Dell’s XPS N1X was embargoed for a May 31 reveal, suggesting it may be one of the earliest products to ship. Pricing has not been confirmed by any OEM.
Is the N1X good for running local AI without the cloud?
Yes — this is arguably its most transformative capability. The full CUDA software stack means tools like Ollama, PyTorch, AUTOMATIC1111, and ComfyUI — the most widely used local AI tools — will run natively on N1X laptops. Currently, running these tools at speed requires either a cloud GPU or a heavy gaming laptop. An N1X-powered thin laptop with RTX 5070-class GPU performance would be the most capable portable local AI machine on Windows ever made.
What is the difference between the N1 and N1X?
Based on leaks: the N1X is the high-performance variant targeting gaming laptops and professional workstations — featuring the full 20-core CPU, 6,144 CUDA cores, and higher power envelope (evidenced by the 245W Lenovo Legion charger). The N1 is the mainstream variant targeting thinner productivity laptops like the ASUS ProArt and Lenovo IdeaPad Slim — likely with a lower power target and reduced GPU core count for better battery efficiency. NVIDIA has not officially confirmed the differentiation between the two variants.
✅ Article Updated — June 1, 2026

This article has been updated following Jensen Huang’s Computex 2026 keynote. RTX Spark confirmed as official consumer brand. HP OmniBook X 14 and OmniBook Ultra 16 added as confirmed OEM devices. Specific device names added for all OEM partners. Jensen Huang quotes added verbatim. 128GB unified memory confirmed. No discrete GPU (dGPU) support confirmed by Tom’s Hardware. Satya Nadella joint keynote June 2 added. More than 30 laptops and 10 desktops planned confirmed by NVIDIA.

⚠️ No Discrete GPU Support — Confirmed

Tom’s Hardware confirms RTX Spark will not support a discrete GPU (dGPU). NVIDIA’s focus with RTX Spark is agentic AI workloads and mainstream gaming — not high-end gaming. RTX 5070-class graphics are integrated within the SoC. Gamers expecting to pair an RTX Spark laptop with an external GPU should note this limitation before purchasing.

Mr Wangdoo — Founder, Wangdoo.com
Mr Wangdoo
Founder & Editor-in-Chief

Mr Wangdoo is the founder and editorial lead of Wangdoo.com, an independent technology news publication launched in September 2025. All articles are fact-checked against primary sources before publication. Where commercial relationships exist, they are disclosed clearly within the relevant article.