Nvidia crashed the Windows PC processor party Intel and AMD have jointly hosted for four decades on Monday, as CEO Jensen Huang used his Computex 2026 keynote in Taipei to unveil the company's first dedicated PC processor. The RTX Spark Superchip, a system-on-chip co-developed with Microsoft and Taiwan's MediaTek, combines a Blackwell GPU with an Arm-based CPU and up to 128GB of unified memory, and it is heading to consumer devices from six major Windows OEMs this fall. Markets responded without ambiguity: Nvidia climbed roughly 4% at the open while Intel shed about 6% and AMD dropped around 5%, each marking a PC chip duopoly that dates to the IBM-compatible era suddenly under threat.
For anyone evaluating a premium Windows laptop or compact desktop purchase before the 2026 holiday season, the Computex announcement redraws the available options entirely. A buyer who purchases an Intel- or AMD-based premium laptop today may find it competing with Nvidia-powered devices targeting the same price bracket by November.
RTX Spark Architecture: What the N1X Processor Delivers
At the heart of the platform is the N1X processor, an Arm-based chip built alongside Taiwan's MediaTek and manufactured on TSMC's 3-nanometer process. At full specification, the N1X pairs 20 Arm CPU cores — 10 performance-class Cortex-X925 cores and 10 efficiency Cortex-A725 cores — with a Blackwell GPU carrying 6,144 CUDA cores, the same count as a desktop RTX 5070. The two dies connect via Nvidia's NVLink C2C chip-to-chip interconnect at up to 300 GB/s of shared bandwidth. The unified memory pool reaches 128GB of LPDDR5X, and the chip targets 1 petaflop of FP4 AI compute — sufficient, Nvidia claims, to run 120-billion-parameter large language models entirely on-device without a cloud subscription.
The key differentiator from Qualcomm's Snapdragon X Elite and Intel's upcoming Panther Lake is not the GPU compute headroom alone but the software ecosystem beneath it. CUDA-X, Nvidia's parallel computing platform, has powered AI research and development for close to two decades. RTX features such as DLSS 4.5 — the company's AI-based frame-generation technology — are confirmed to run natively on RTX Spark hardware, and Adobe has committed to native Arm builds of both Photoshop and Premiere Pro for the platform — a concession Qualcomm's Snapdragon X could not secure after two years on the market.
One structural limitation is worth flagging at launch: RTX Spark does not support a discrete external GPU, meaning the Blackwell GPU integrated into the SoC is the ceiling for graphics performance. That positions the platform primarily around AI agents, creativity workflows, and mainstream gaming rather than enthusiast-level frame rates.
Read more: Nvidia RTX Spark Superchip: Windows PC Chip With Full CUDA Stack Targets Dell, Microsoft This Fall
No Benchmarks Yet: What Intel Said and What Analysts Warned
Nvidia provided no comparative benchmarks against rival chips at the keynote. ServeTheHome's analysis classified the announcement as closer to a teaser than a technical briefing, noting that Nvidia did not address how RTX Spark compares to rival chips from Intel, AMD, or Qualcomm. The company said it will release performance metrics closer to market launch.
Intel responded on the record. Nish Neelalojanan, Senior Director of Product Management for Intel's Client Computing Group, told Tom's Hardware that the company approaches Nvidia's PC market entry with what he called "a healthy dose of paranoia," while flagging that Windows on Arm platforms still face unresolved app compatibility and digital rights management challenges that x86 systems do not. Qualcomm SVP Kedar Kondap took a more relaxed position, welcoming Nvidia's entry as an endorsement of the broader Arm ecosystem his company spent years building.
DigiTimes analyst Jason Tsai warned that RTX Spark could remain a niche luxury product unless complete systems land near the $1,500 price point. Nvidia has not announced pricing; the company confirmed only that initial devices will target the premium segment.
Six Major Windows OEMs, Over 30 Laptops, 10 Desktops
Confirmed launch partners include Microsoft, Dell, HP, ASUS, Lenovo, and MSI. Microsoft simultaneously revealed the Surface Laptop Ultra, a 15-inch device built on RTX Spark from the silicon up, with a mini-LED PixelSense Ultra display and 128GB of unified memory. Microsoft Surface product leader Andrew Hill described it as the most performance-oriented Surface the company has ever built.
Nvidia confirmed that over 30 laptops and 10 desktops powered by RTX Spark will reach shelves this fall, with the thinnest designs measuring 14 millimeters. On supply, Nvidia said it does not expect RTX Spark device availability to be constrained despite ongoing global memory price pressure — a pointed reassurance given that DRAM costs have pushed projected PC prices higher across the industry this year.
Vera Rubin Enters Full Production: Data Centers Next
Alongside the PC announcement, Huang confirmed that Nvidia's next-generation data center AI platform — Vera Rubin — has entered full production, with first systems expected to ship this fall. The platform pairs Nvidia's Vera CPUs with Rubin GPUs over sixth-generation NVLink, delivering approximately 3.5 times the AI training performance and five times the inference performance of its Blackwell predecessor.
Early customers include Anthropic, OpenAI, SpaceX's xAI, Dell, Oracle, and CoreWeave. The CoreWeave delivery was underscored in real time when Dell CEO Michael Dell posted on X during the keynote that his company had delivered its AI servers — built on Vera Rubin NVL72 racks — to the cloud infrastructure provider.
Huang framed the CPU push as a response to a structural shift in AI workloads: inference and agentic AI place far higher demand on general-purpose CPU cores than model training did, creating headroom for Nvidia to enter a market Intel and AMD have owned in data centers.
Arm Holdings Surges 15%: Why the Architecture Story Matters
The clearest single-day beneficiary was Arm Holdings, whose shares rose roughly 15% in the US session — because every Arm-based laptop and data center CPU that ships, including Nvidia's N1X, pays architecture royalties to Arm. Qualcomm fell roughly 6–7%, reflecting added competitive pressure on its Snapdragon X franchise from a company whose software ecosystem already commands deep loyalty among developers and enterprises.
Nvidia's own year-to-date trajectory heading into Computex had been comparatively restrained — up roughly 13% against the iShares Semiconductor ETF's 89% rise — which contributed to the strength of the market's reaction. Investors had not fully priced in a credible Nvidia bid for consumer PC silicon on top of its data center dominance. Both AMD and Intel remain up sharply for the year despite Monday's declines — Intel by nearly 200%, AMD by roughly 130% — making any signal of renewed competition at the high end of the Windows PC market an outsized sensitivity.
What Nvidia brings to the consumer market is not merely a chip but a software stack, an OEM relationship set, and a brand synonymous with AI performance. Those are three assets Intel and Qualcomm have spent years trying to assemble in the PC space and AMD has not yet fully entered. That asymmetry is what sent Intel and AMD lower and Arm Holdings higher on the same morning.
Computex 2026 runs through June 5, with additional announcements expected from AMD, Intel, and Qualcomm throughout the week. The fall hardware cycle will deliver the decisive test: whether consumer demand for on-device AI agents — and the premium prices the RTX Spark platform will command — can translate Monday's keynote momentum into a new and durable revenue line for Nvidia.
Frequently Asked Questions
What is the Nvidia RTX Spark chip?
RTX Spark is Nvidia's first PC processor, a system-on-chip pairing a Blackwell GPU with an Arm-based N1X CPU co-developed with Microsoft and MediaTek. It targets Windows laptops and compact desktops, supports up to 128GB of unified memory, and delivers 1 petaflop of FP4 AI compute — enough to run large language models entirely on-device without a cloud connection.
When do RTX Spark laptops go on sale?
Nvidia and its six confirmed OEM partners — Microsoft, Dell, HP, ASUS, Lenovo, and MSI — are targeting a fall 2026 launch. Over 30 laptop models and 10 desktop configurations are expected by the holiday season. No pricing has been announced, though Nvidia says supply will not be limited.
How does RTX Spark compare to Intel and AMD chips?
Nvidia has not released comparative benchmarks, stating it will do so closer to the retail launch. The Blackwell GPU inside RTX Spark is described as roughly equivalent to the desktop RTX 5070. Intel's senior product director for client computing acknowledged the competitive pressure while flagging that Windows on Arm still faces app compatibility and DRM challenges that x86-based chips do not.
Why did Intel and AMD stock fall on the RTX Spark news?
Nvidia's entry into Windows PC processors — a segment Intel and AMD have jointly controlled for decades — signals new competition at the premium end of the market. Intel fell roughly 6% and AMD dropped around 5% at Monday's open. Both companies remain up sharply year-to-date (Intel ~200%, AMD ~130%), but the RTX Spark announcement raises questions about their ability to defend margins in high-end Windows PCs once Nvidia devices reach shelves this fall.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.
