Nvidia has locked in October 16, 2026 as the launch date for RTX Spark, the Windows PC platform built around its N1X chip, a date Nvidia itself has now confirmed and which is echoed in reports from The Verge and Wccftech published September 3, 2026. The chip will ship in two configurations across laptops and compact desktops from six named launch-day manufacturers, with PC Gamer now reporting launch pricing spanning roughly $2,600 to $7,000 depending on configuration, marking Nvidia’s first serious attempt to sell a full CPU-and-GPU platform directly into the Windows PC market it has never owned outright.
The move caps roughly four months of staged reveals that began with Jensen Huang’s Computex 2026 keynote on May 31, 2026, continued through additional spec disclosures over the summer, and ended with prototype benchmark leaks just ahead of the confirmed October date. With a firm October ship date now confirmed, the RTX Spark N1X becomes one of the most closely watched hardware launches of the year for anyone tracking the AI PC category, and it carries real stakes for Intel, AMD, and Qualcomm’s competing Arm and x86 platforms.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
Nvidia Confirms October Launch for RTX Spark N1X
The Verge reported that Nvidia’s N1X chip is arriving in RTX Spark laptops in two different configurations in October 2026, ending months of speculation about exact timing. Wccftech’s September 3 report echoed the same window, noting that the top-tier N1X configuration lands in both laptops and compact desktops at launch, while the second-tier configuration is laptop-only when it first ships.
This is a meaningfully fast turnaround by Nvidia’s standards. The company only unveiled the RTX Spark superchip publicly at its Computex 2026 keynote on May 31, 2026, and spent the following weeks fielding questions about whether the platform could run existing Windows 11 software at all. About four months later, Nvidia has moved from concept reveal to a dated retail launch with named OEM partners already lined up, according to Tom’s Hardware’s Computex coverage.
Nvidia RTX Spark systems are being framed by the company not as a niche AI accessory but as a mainstream Windows PC category, positioned to compete for shelf space against conventional laptops built around Intel, AMD, and Qualcomm silicon. That framing explains why the October date matters: it puts RTX Spark hardware in front of holiday-season shoppers and enterprise buyers finalizing 2027 procurement budgets at the same time.
Inside the N1X Chip: Two Configurations Explained
Wccftech’s September 3 breakdown lays out the two N1X configurations in detail. The top-end design pairs a 20-core Nvidia Grace CPU with a Blackwell RTX GPU carrying 6,144 CUDA cores, a core count Wccftech noted matches the desktop RTX 5070. That configuration supports between 24GB and 128GB of unified memory. The second, lower configuration steps down to an 18-core Grace CPU and a Blackwell RTX GPU with 5,120 CUDA cores, which Wccftech pointed out is still more cores than found in the mobile RTX 5070, paired with 24GB or 32GB of unified memory, figures Windows Central independently corroborated in its own September 2026 reporting on the base N1X configuration.
| Spec | Configuration 1 (Top) | Configuration 2 (Base) |
|---|---|---|
| CPU | 20-core Nvidia Grace CPU | 18-core Nvidia Grace CPU |
| GPU | Blackwell RTX GPU, 6,144 CUDA cores | Blackwell RTX GPU, 5,120 CUDA cores |
| Unified memory | 24GB – 128GB LPDDR5X | 24GB – 32GB LPDDR5X |
| Memory bandwidth | Up to 300 GB/s | Not disclosed |
| Process node | TSMC 3nm | TSMC 3nm |
| TDP | Up to 80W | Up to 80W |
| I/O | USB4, Thunderbolt | USB4, Thunderbolt |
| Device availability at launch | Laptops and compact desktops | Laptops only |
Those figures come from a combination of Wccftech’s launch report and Tom’s Guide’s spec summary, which additionally reported the platform runs on a TSMC 3nm process, supports USB4 and Thunderbolt connectivity, and tops out around 80 watts of TDP in laptop form, though a Wikipedia entry updated in September 2026 puts the compact desktop variant’s thermal envelope closer to 100 watts. Nvidia’s own launch materials, as relayed by Tom’s Hardware and PCWorld, claimed the platform delivers up to one petaflop of FP4 AI compute at its top-end configuration, though Nvidia has not published independent third-party validation of that figure.
From GB10 to N1X: How Nvidia Repackaged DGX Spark Silicon
The RTX Spark N1X did not appear out of nowhere. Tom’s Hardware’s Computex wrap-up described the chip as the Windows-PC version of silicon long referred to internally as N1X, while the closely related product Nvidia already sells to AI developers uses the GB10 branding inside the DGX Spark mini workstation. Nvidia’s DGX Spark, reviewed by Tom’s Hardware back in January 2026, ships with 128GB of RAM and an RTX 5070-class Blackwell GPU paired with what Tom’s Hardware described as a Mediatek-designed 20-core Arm CPU complex, but runs a Linux-based operating system rather than Windows.
The Register’s June 1, 2026 report on the announcement made the lineage explicit, noting that the same Grace Blackwell superchip design powering Nvidia’s DGX Spark AI workstations is now heading to Windows PCs under the RTX Spark name. In effect, Nvidia took hardware it originally built for AI researchers running local large language models and repackaged it, with a different operating system and branding, for general Windows PC buyers.
| Spec | RTX Spark N1X (Config 1) | DGX Spark / GB10 |
|---|---|---|
| CPU | 20-core Nvidia Grace CPU | Mediatek-designed 20-core Arm CPU complex |
| GPU | Blackwell RTX, 6,144 CUDA cores | RTX 5070-class Blackwell GPU |
| Memory | 24GB – 128GB unified LPDDR5X | 128GB LPDDR5X |
| Operating system | Windows 11 (Windows on Arm) | Linux-based DGX OS |
| Primary use case | Consumer/creator laptops, mini PCs | AI developer workstation |
| Launch status | Launching October 2026 | Already shipping (reviewed January 2026) |
That shared heritage is why so much of the pre-launch conversation about Nvidia RTX Spark has centered on AI workloads rather than traditional PC benchmarks. A 128GB unified memory pool is unusual for a consumer laptop chip, and Tom’s Hardware reported it exists specifically to give large AI models, including some in the 120-billion-parameter range, enough headroom to run locally rather than through a cloud API.
Which Laptop and Desktop Makers Are Building RTX Spark Systems
Nvidia’s own newsroom post, published alongside the Computex reveal, named ASUS, Dell, HP, Lenovo, Microsoft, and MSI as the six launch partners building RTX Spark laptops and compact desktops, with Nvidia separately confirming that same June 2026 that Acer and Gigabyte models would follow after the initial wave. That is a broad OEM roster for a first-generation platform, spanning both PC-industry incumbents and Microsoft’s own Surface line, and it signals that Nvidia secured buy-in across the Windows ecosystem rather than launching with a single reference design.
Some of those same partners have already been active in the RTX Spark story ahead of the October launch. Our earlier coverage detailed how ASUS and MSI RTX Spark laptop allocations sold out during earlier limited availability windows, and how ASUS moved to reorder RTX Spark stock as eight OEMs positioned inventory for the fourth quarter. That pattern of early sellouts ahead of a formal wide launch suggests Nvidia and its partners are anticipating strong initial demand once October retail availability actually opens.
The mix of laptop and compact desktop form factors is also notable. Wccftech’s reporting specified that only the top-tier N1X configuration is confirmed for compact desktops at launch, meaning shoppers who want a Spark-branded mini PC rather than a laptop will likely be steered toward the pricier 128GB-capable variant rather than the base configuration. That desktop variant also appears to run hotter in practice: the September 2026 Wikipedia update on the platform cites a roughly 100-watt thermal envelope for compact desktop units, above the approximately 80-watt figure reported for the laptop configurations.
Windows on Arm: Betting on Agentic AI
RTX Spark runs Windows 11 on Arm rather than the Linux environment used by DGX Spark, a distinction both The Verge and The Register emphasized in their coverage. Tom’s Hardware described Nvidia and Microsoft’s stated goal as turning Windows into what the companies call an agentic AI operating system, with the RTX Spark hardware providing the local compute headroom needed to run AI agents and large models without relying entirely on cloud infrastructure.
Nvidia’s marketing language for the launch, as reported by The Verge and Tom’s Hardware, describes RTX Spark as “the most efficient PC chip ever built,” and frames the broader push as an effort to “reinvent” Windows PCs for the AI-agent era. Those are Nvidia’s own characterizations rather than independently verified claims, and neither outlet reported third-party efficiency benchmarks confirming the superlative at time of writing.
The Windows-on-Arm choice carries real risk. Windows Latest reported on June 2, 2026 that Nvidia said the platform would run all Windows 11 apps, but the outlet also noted Nvidia had not yet shared CPU benchmarks or other technical details at that point, leaving open questions about real-world application compatibility and performance on Arm-native Windows, a platform that has struggled with software compatibility in prior Qualcomm-based attempts.
RTX Spark vs Apple Silicon: The MacBook Comparison
Much of the press framing around RTX Spark N1X has drawn direct lines to Apple’s MacBook Pro lineup. Tom’s Guide’s price-prediction piece, published under the headline referencing whether “Apple silicon should be terrified,” argued that the 128GB unified memory ceiling and integrated CPU-GPU design put RTX Spark in a similar conceptual bracket to Apple’s own unified-memory Mac chips, even though the two platforms use entirely different architectures and operating systems.
Tom’s Guide followed up on August 11, 2026 with a report on leaked Geekbench scores from an RTX Spark test unit, again framing the results around the question of whether Apple should be concerned about what the outlet called potentially “the fastest Windows laptops on the planet.” The outlet did not publish specific benchmark figures alongside that framing in the portion of its coverage available at time of writing, so any direct numeric comparison to Apple’s current chips remains unconfirmed pending fuller third-party testing.
The comparison extends beyond Apple’s hardware and into its market position more broadly. Nvidia’s own compute ambitions have already reshaped demand elsewhere in the Apple ecosystem this year, as detailed in our report on OpenAI buying Mac hardware around the same period RTX Spark units sold out, illustrating how tightly the AI-hardware buying cycle now links Nvidia’s roadmap to Apple’s supply chain even when the two companies are nominally competitors.
RTX Spark vs Qualcomm and AMD: The Competitive Landscape
Beyond Apple, RTX Spark lands squarely in territory Qualcomm has tried to claim with its Snapdragon X Arm-based Windows chips, and where AMD has pushed its Ryzen AI processors as the go-to option for on-device AI features. Tom’s Hardware’s Computex wrap-up noted the RTX Spark reveal came alongside Intel’s own Xeon 6+ announcement at the same event, and our prior coverage detailed how Intel’s Xeon 6+ launch was overshadowed by the RTX Spark reveal in the press cycle that week.
That overshadowing effect has continued to play out in financial markets since the initial announcement. Our earlier reporting on how Nvidia’s RTX Spark reveal weighed on Intel and AMD shares captured investor concern that Nvidia entering the Windows PC chip market directly threatens both companies’ existing laptop silicon businesses, since RTX Spark competes for the same OEM design wins that traditionally went to Intel or AMD processors.
Qualcomm faces a different kind of pressure. Its Snapdragon X platform helped popularize Windows on Arm as a viable category after years of false starts, but RTX Spark now offers OEMs an Arm-based Windows alternative backed by Nvidia’s GPU and AI software ecosystem, a combination Qualcomm cannot currently match on the graphics and CUDA-software side.
Early Benchmark Leaks: What Cinebench and Geekbench Show So Far
Two rounds of pre-launch benchmark leaks have already circulated. Wccftech reported on July 28, 2026 that an early RTX Spark prototype produced Cinebench 2026 scores of 540 points single-core and 5,771 points multi-core, while cautioning in its own headline that the early prototype “leaves plenty of performance on the table,” suggesting the retail silicon and drivers should improve on those numbers by the October launch.
Tom’s Guide’s August 11, 2026 report covered a second leak, this time Geekbench scores from what it described as a more representative test unit, again raising the Apple-comparison framing without publishing exhaustive score breakdowns in the portion of coverage cited here. Those Geekbench 6 figures did circulate separately: listings surfaced by Notebookcheck and VideoCardz on August 7 and 8, 2026 put the top 20-core N1X engineering sample at roughly 2,570 single-core and 23,126 multi-core, with the 18-core configuration landing near 2,541 single-core and 21,776 multi-core. Both outlets stressed the numbers came from engineering samples rather than final retail hardware, so the shipping silicon and drivers could still move those figures before the October launch. Combined, the two leaks indicate independent testers already have access to pre-release RTX Spark hardware roughly two months ahead of the confirmed October launch, which is typical for a major PC platform reveal of this scale.
PCWorld’s coverage added a gaming-specific data point to the pre-launch narrative, running under a headline promising “100fps gaming promises” for RTX Spark systems, reflecting Nvidia’s own marketing claims about gaming performance on the platform rather than independently benchmarked results. Given that RTX Spark’s GPU core counts sit below current discrete desktop RTX 50-series cards, real-world gaming performance at launch will likely depend heavily on driver optimization and game-specific support for the integrated Blackwell GPU architecture.
Pricing: What Reports Say So Far
Nvidia and its OEM partners have now moved past speculation on pricing. With RTX Spark N1X laptops and compact desktops confirmed to ship October 16, 2026, PC Gamer reported that retail pricing across the lineup spans roughly $2,600 to $7,000, a far wider spread than Tom’s Guide’s earlier prediction piece anticipated when it was estimating a likely range based on the chip’s specifications and Nvidia’s existing DGX Spark pricing precedent. That $2,600–$7,000 band now functions as the confirmed reference point for shoppers, superseding the analyst-style estimates that circulated before OEMs finalized their own laptop and desktop SKUs.
The two configurations do command different price points given their memory ceilings, and Microsoft’s own Surface lineup now illustrates the spread concretely. Microsoft’s Surface Laptop Ultra starts at $2,599 with 24GB of RAM at the base end, according to Microsoft’s own listing, while the company’s RTX Spark Dev Box, aimed squarely at AI developers, runs $5,999 for a 128GB configuration, per PCMag’s coverage. That roughly $3,400 gap between Microsoft’s entry-level and top-end RTX Spark machines mirrors the memory-tier split across ASUS, Dell, HP, Lenovo, and MSI SKUs now going live around the October 16 launch window.
Market Impact: Nvidia’s Direct Push Into Windows PCs
The scale of the market Nvidia is entering is significant on its own terms. RTX Spark represents the company’s first attempt to sell a complete CPU-and-GPU platform directly into the mainstream Windows laptop and desktop category, a market historically dominated by Intel and AMD processors paired with Nvidia’s own discrete GPUs as an add-on component rather than an integrated Nvidia-designed system.
That shift changes Nvidia’s relationship with its own long-time OEM partners. Instead of simply supplying discrete GPUs that OEMs pair with Intel or AMD processors, Nvidia is now asking those same OEMs, ASUS, Dell, HP, Lenovo, MSI, and Microsoft itself through Surface, to build systems around a complete Nvidia-designed CPU-GPU platform. Early stock movements documented in our coverage of the reveal’s effect on Intel and AMD shares suggest investors are already pricing in some disruption to the traditional PC-chip supply relationship.
Historical Context: From AI Workstation to Windows PC
The RTX Spark story traces back to Nvidia’s DGX Spark, reviewed favorably by Tom’s Hardware in January 2026 as a well-rounded local AI toolkit that outperformed AMD’s Ryzen AI Max+ 395 in the outlet’s testing, but which ran Linux rather than Windows and targeted AI developers rather than mainstream PC buyers. Nvidia and Microsoft then teased “a new era of PC” through coordinated social media posts ahead of Computex 2026, according to Tom’s Hardware’s late-May reporting, before formally unveiling RTX Spark and the N1X branding on May 31, 2026.
That five-month arc, from a DGX Spark review in January, to a Windows-focused reveal in May, to a dated October retail launch, represents an unusually fast repositioning of a single piece of silicon across two entirely different operating systems and buyer categories. It also reflects how central local AI compute has become to Nvidia’s product strategy across every tier of its lineup, from data-center GPUs down to what is now a consumer Windows laptop chip.
Nvidia’s Roadmap Beyond the October Launch
Nvidia has signaled RTX Spark is not a one-off product. Tom’s Hardware’s Computex coverage reported that Jensen Huang promised every future generation of Nvidia’s platforms will include a Spark chip, with a roadmap running through the Rubin architecture generation followed by chips code-named Rosa and then Feynman. That framing positions the October 2026 N1X launch as the first entry in what Nvidia intends to be an annual or near-annual cadence of Spark-branded hardware, rather than a single experimental product testing Windows PC market appetite.
For OEM partners and enterprise buyers, that multi-generation commitment matters as much as the specs of the first chip. A company deciding whether to build a product line around RTX Spark silicon is effectively betting on Nvidia sustaining Windows PC chip development for several years, following the same pattern that made its data-center GPU roadmap a reliable planning anchor for cloud providers.
What This Means for AI PC Adoption in 2026
The AI PC category, part of the broader shift in AI chip development we track closely, has struggled to define itself clearly since Intel, AMD, and Qualcomm all began marketing on-device neural processing units in 2024 and 2025 without a signature workload that clearly needed the extra silicon. RTX Spark’s 128GB unified memory pool gives Nvidia a much more concrete pitch: running large local AI models, including some at the 120-billion-parameter scale, directly on the device rather than through a cloud API, according to Tom’s Hardware’s specification reporting.
Whether mainstream buyers actually want to run 120-billion-parameter models locally, versus simply using cloud-based AI assistants through a browser or app, remains an open question the October launch will start to answer. Enterprise IT buyers evaluating data privacy or offline-capability requirements are the most obvious early audience for that specific pitch, ahead of general consumer adoption.
Predictions: What to Watch Before and After October
- Independent Cinebench and Geekbench benchmarks from retail RTX Spark units should surface within days of the October ship date, building on the prototype leaks Wccftech and Tom’s Guide already reported over the summer.
- Official pricing from ASUS, Dell, HP, Lenovo, Microsoft Surface, and MSI will likely be staggered by configuration, with base 32GB units priced well below top 128GB-capable models once OEMs finalize SKUs.
- Expect continued Apple-comparison coverage to intensify once real benchmark data replaces the current speculation, particularly around on-device AI model performance rather than traditional CPU or gaming benchmarks.
- Compact desktop availability will likely stay limited to the top N1X configuration through the initial launch window, with the base configuration remaining laptop-exclusive into 2027 based on current OEM signaling.
- Given Nvidia’s stated Rubin-Rosa-Feynman roadmap for future Spark chips, the October N1X launch should be treated as the opening entry in a recurring annual hardware cycle rather than an isolated product experiment.
These are forward-looking assessments based on Nvidia’s confirmed roadmap statements and OEM launch patterns reported by named outlets, not guaranteed outcomes, and actual October pricing, availability, and benchmark results could shift these expectations in either direction.
Frequently Asked Questions
When does the Nvidia RTX Spark N1X launch?
Nvidia’s RTX Spark N1X is set to launch on October 16, 2026 for laptops and compact desktops, according to The Verge and Wccftech reporting published September 3, 2026, with PC Gamer confirming the ship date and initial pricing ahead of release.
What is the difference between the two N1X configurations?
The top configuration pairs a 20-core Grace CPU with a 6,144-core Blackwell RTX GPU and up to 128GB of unified memory, available in both laptops and compact desktops. The base configuration uses an 18-core Grace CPU with a 5,120-core Blackwell RTX GPU and up to 32GB of unified memory, and is laptop-only at launch.
Is RTX Spark the same chip as the DGX Spark GB10?
They share the same underlying Grace Blackwell superchip design, but RTX Spark runs Windows 11 on Arm and targets mainstream PC buyers, while DGX Spark uses the GB10 branding, runs a Linux-based operating system, and targets AI developers.
Which laptop and desktop brands will sell RTX Spark systems?
Nvidia’s newsroom confirmed ASUS, Dell, HP, Lenovo, Microsoft Surface, and MSI as launch partners, with Acer and Gigabyte models following afterward.
How much does the Nvidia RTX Spark N1X cost?
Pricing is now confirmed: PC Gamer reports RTX Spark laptops and compact desktops span roughly $2,600 to $7,000 at the October 16, 2026 launch, with Microsoft’s Surface Laptop Ultra starting at $2,599 (24GB RAM) and its higher-end RTX Spark Dev Box priced at $5,999 (128GB RAM), per PCMag.
Will RTX Spark run all Windows 11 apps and games?
Nvidia told Windows Latest in June 2026 that RTX Spark would run all Windows 11 apps, though the outlet noted Nvidia had not yet shared full CPU benchmarks or compatibility details at that time. Broader real-world compatibility testing is expected once retail units ship in October.
How does RTX Spark compare to Apple Silicon?
Tom’s Guide has framed RTX Spark’s unified memory design and AI compute focus as a direct challenge to Apple’s MacBook Pro lineup, though the two platforms use different architectures and operating systems, and detailed independent benchmark comparisons were not yet published at the time of this report.
What operating system does RTX Spark use?
RTX Spark runs Windows 11 on Arm, distinguishing it from the Linux-based DGX Spark that uses the related GB10 chip.
Does RTX Spark replace Nvidia’s discrete GPU business?
No. RTX Spark is a separate integrated CPU-GPU platform aimed at the Windows laptop and mini PC market. Nvidia continues to sell discrete RTX GPUs separately for desktop gaming and workstation builds.
