Nvidia RTX Spark Superchip: A Complete Guide to the AI Powerhouse Reshaping Windows PCs

Jensen Huang took the stage at GTC Taipei on May 31, 2026, and announced a chip that could fundamentally redraw the consumer PC landscape. The Nvidia RTX Spark superchip, codenamed N1X, is the company's first Arm-based system-on-chip designed for Windows laptops and compact desktops . Co-developed with MediaTek and in tight collaboration with Microsoft, the chip is built to power a new class of "personal AI agent" machines that can run massive 120-billion-parameter models locally, eliminating the need for cloud round-trips .

What Is the RTX Spark Superchip?

RTX Spark is a single package that fuses two chiplets—a Blackwell RTX GPU and a 20-core Grace CPU—over Nvidia’s NVLink-C2C interconnect at 600 GB/s bidirectional bandwidth . It is manufactured on TSMC’s advanced 3nm process and contains roughly 70 billion transistors . The result is a Windows on Arm platform that delivers up to 1 petaflop of AI compute (FP4 with sparsity) and supports up to 128 GB of unified LPDDR5X memory, shared seamlessly between the CPU and GPU .

Unlike previous Nvidia hardware for PCs, RTX Spark isn’t a discrete GPU you add to an Intel or AMD system—it is the primary processor. It runs the full Nvidia AI software stack, including CUDA, TensorRT, NeMo, and NIM microservices, making it the first Windows laptop chip to offer native CUDA acceleration .

Key Specifications

• GPU: Blackwell RTX with 6,144 CUDA cores and 5th-generation Tensor Cores (FP4)
• CPU: 20-core Arm-based Grace CPU (10 Cortex-X925 performance cores + 10 Cortex-A725 efficiency cores), co-developed with MediaTek
• Process node: TSMC 3nm, approximately 70 billion transistors
• Unified memory: Up to 128 GB LPDDR5X with up to 300 GB/s bandwidth
• AI performance: 1 petaflop (FP4); up to 1,000 TOPS inference
• Interconnect: NVLink-C2C, 600 GB/s bidirectional bandwidth
• GPU performance target: Comparable to a GeForce RTX 5070 Laptop GPU but with significantly lower power draw

RTX Spark Laptops and Desktop Lineup

Microsoft confirmed that RTX Spark devices will ship in fall 2026 from six major OEMs, with models from Acer and Gigabyte expected to follow . Nvidia indicated that roughly 30 laptop models and 10 desktop models are in the pipeline .

Confirmed RTX Spark devices include:

• ASUS ProArt P16 and ProArt P14: Creator-focused 16-inch and 14-inch laptops with Lumina Pro OLED displays
• Dell XPS 16: A premium thin-and-light targeting power users
• HP OmniBook X 14 and OmniBook Ultra 16: AI-native consumer and prosumer notebooks
• Lenovo Yoga Pro 9n: A performance-oriented convertible for creative work
• Microsoft Surface Laptop Ultra: The flagship Windows on Arm device, positioned as a direct MacBook Pro competitor
• MSI Prestige N16 Flip AI+: The only announced 2-in-1 RTX Spark laptop

Early reports note that devices can be as thin as 14 mm and weigh around three pounds, a form factor made possible by the integration of CPU, GPU, and memory into a single package .

Market Reaction: Intel, AMD, and Qualcomm Take a Hit

The announcement triggered an immediate selloff across the PC chip sector. Qualcomm (QCOM) suffered the sharpest decline, as RTX Spark directly challenges its Snapdragon X exclusivity on Windows on Arm devices. The stock fell as much as 10% in pre-market trading and closed down approximately 7%–8.77%, wiping out over $10 billion in market capitalization .

Intel and AMD were also punished. Intel shares dropped around 6%, while AMD fell approximately 5%, as investors priced in a long-term disruption risk to the x86 duopoly that has defined the PC industry for decades . Nvidia’s own stock rose roughly 4–5%, ending near $222, while AI-exposed names like Arm Holdings, ServiceNow, and Micron gained ground .

The market reaction reflected a clear verdict: RTX Spark expands Nvidia’s AI dominance off the server rack and directly into the $200 billion client PC market .

Qualcomm’s Response: “Welcome to the Family”

Qualcomm SVP of Computing Kedar Kondap addressed the new competition directly during a media Q&A at Computex. “Welcome to the family—we’re excited,” he said, framing Nvidia’s entry as a validation of the Arm-based Windows ecosystem that Qualcomm had been building for years .

Kondap noted that the real competition remains the x86 entrenched base, not fellow Arm challengers, and pointed to Qualcomm’s newly announced Snapdragon C platform as the company’s answer for lower price points—a move that effectively concedes the high-end AI performance crown to Nvidia while protecting volume and efficiency segments .

Qualcomm CEO Cristiano Amon separately told investors the company can absorb the competition and still reach its $22 billion non-handset revenue target by fiscal 2029 .

How RTX Spark Compares to Snapdragon X Elite

No independent benchmarks exist yet, but the architectural gulf is enormous:

• Snapdragon X Elite: 12 Oryon CPU cores, up to 45 TOPS NPU, no discrete-class GPU. A monolithic SoC designed for efficiency and moderate AI workloads .
• Snapdragon X2 Elite (upcoming): Expected CPU and NPU improvements but remains a CPU-centric design with integrated graphics—no RTX-class GPU .
• RTX Spark: Brings a full Blackwell RTX GPU with 6,144 CUDA cores, ray tracing, and Tensor Cores delivering 1 petaflop of AI compute—roughly 20x the AI TOPS of the X Elite’s NPU. With 128 GB unified memory and NVLink-C2C, it can run local 120B-parameter models with million-token context, workloads that current Snapdragon chips cannot touch .

Snapdragon X will remain more power efficient for light CPU tasks and presumably cheaper, but RTX Spark radically raises the performance ceiling for Arm-based Windows laptops—something neither Qualcomm nor Apple has offered .

Expected Price Range

Morgan Stanley estimates cited in press reports put RTX Spark machines in the $1,799–$2,899 range . Some early retailer leaks suggest entry configurations could start around $1,499, with fully loaded Surface Laptop Ultra bundles pushing above $2,800 . In all cases, RTX Spark occupies a premium tier comparable to Apple’s MacBook Pro lineup—not budget territory.

What This Means for the Arm-Based Windows PC Market

1. The Qualcomm exclusivity era is over. Until this announcement, Snapdragon X was the only Windows on Arm silicon available. RTX Spark introduces credible GPU-driven competition that fundamentally changes the calculus for OEMs and customers .

2. Massive AI performance leap for Arm laptops. With 1 petaflop of on-device AI compute and 128 GB unified memory, RTX Spark machines can run 120B-parameter models locally—narrowing Apple’s Mac advantage and unlocking entirely new on-device AI agent workloads .

3. Direct threat to Intel and AMD. Nvidia’s entry with an Arm-based chip threatens the x86 duopoly that Intel and AMD have controlled for four decades. Both companies saw their shares fall sharply on the announcement . Intel product executives publicly acknowledged the competitive risk with what one called “a healthy dose of paranoia” .

4. Microsoft’s Arm bet reaches critical mass. Microsoft co-announced RTX Spark and revealed the Surface Laptop Ultra as the flagship device. This signals that Windows on Arm is no longer an experiment—it is a strategic priority backed by Nvidia’s silicon .

5. The competitive map is splitting by price tier. Qualcomm is pivoting to its Snapdragon C platform targeting laptops from about $600 downward, while RTX Spark stakes out the $1,799+ premium AI and creative workstation tier. AMD and Intel will increasingly find themselves squeezed from both ends .

6. Apple’s MacBook finally gets a direct Arm challenger. RTX Spark is the first Windows-on-Arm platform to match or exceed Apple Silicon’s unified memory and AI performance, with the added advantage of full CUDA compatibility for creative and machine learning workflows .