Nvidia has locked in October as the month its first RTX Spark PCs reach store shelves, and the launch roster is now public: eight OEMs, led by Lenovo and Acer, are attached to the rollout, according to Reuters and StorageReview. It is the clearest signal yet that Nvidia intends to sell complete Windows machines, not just the silicon inside them, and it puts the company in direct competition with the laptop and desktop brands that have been its customers for three decades.
The RTX Spark laptop and mini PC wave matters because it is the first time Nvidia has put its own branded platform, built around the same Grace Blackwell architecture that powers its DGX Spark developer workstation, into mainstream retail channels. StorageReview reported that Nvidia says RTX Spark PCs are “coming this October,” and Reuters separately confirmed that Lenovo and Acer will launch the first Windows PCs powered by the RTX Spark chip that same month. Six more OEMs are attached to the platform, with some following later in the fall.
Nvidia Confirms October Launch for RTX Spark PCs
The October timing itself is confirmed across multiple outlets, even though nobody has published an exact calendar date. Reuters, StorageReview, Gigazine, TechSpot and Gadgets 360 have all reported the same month-level window: RTX Spark laptops and desktops begin arriving in October 2026. What’s still unconfirmed is which specific week, or whether Nvidia and its partners will stagger the date by region, the way NVIDIA has historically done with GPU launches.
What is confirmed is the shape of the lineup. Lenovo and Acer are the two OEMs named as launching the first Windows PCs built on the RTX Spark chip in October, per Reuters. That is a notably narrow first wave for a platform Nvidia has talked about since Computex 2026, when the company’s own materials described a much broader initial partner list. The gap between that early announcement and the smaller confirmed October cohort is itself a data point worth watching, and it’s one of the threads this piece pulls on below.
What Is RTX Spark, and Why Nvidia Is Building Whole PCs
RTX Spark is the name Nvidia is using for both the chip and the finished PCs built around it, a naming choice that mirrors how Apple treats “M-series” or “Mac.” In its own announcement, Nvidia described the platform in specific terms: “NVIDIA RTX Spark powers the world’s first Windows PCs purpose-built for personal agents, featuring 1 petaflop of AI performance, industry-leading power efficiency, full-stack NVIDIA AI and graphics technology, and up to 128GB of unified memory,” the company said in its May 2026 news release.
That 128GB unified memory figure is the number to focus on. It’s roughly what you’d need to run a mid-sized open-weight language model entirely on-device, without a round trip to a cloud API, which is the entire pitch behind the “personal agents” framing. Nvidia is betting that Windows users will want AI agents running locally on their own hardware, and that the bottleneck to making that useful isn’t raw compute so much as it is having enough fast memory sitting next to the GPU to hold the model weights.
At Computex 2026, Nvidia CEO Jensen Huang framed the ambition even more directly. “This is going to be the new PC,” Huang said from the keynote stage, according to Fortune’s coverage of the event. He went further in the same appearance, telling the audience that “Microsoft and Nvidia are going to reinvent the PC,” a line Fortune also captured. Coming from the head of the company that makes the GPU inside most gaming and workstation PCs already, that’s a claim about market position as much as it is about product design.
The OEM Lineup: Lenovo and Acer Move First
Reuters’ reporting is specific on who ships first: Lenovo and Acer will launch the first Windows PCs powered by the RTX Spark chip in October. StorageReview adds more texture, referencing Lenovo, Acer, and six other OEMs shipping RTX Spark Windows PCs in that window, though that outlet’s excerpt does not fully enumerate all six by name or confirm each one’s exact October shipping status.
That distinction matters for anyone actually shopping in October. If you want an RTX Spark machine the month it launches, the safest bet based on current reporting is a Lenovo or Acer model. Everyone else attached to the platform, including some very large PC brands, is described as following later in the fall rather than hitting shelves in October itself.
Lenovo’s RTX Spark Laptops: Yoga Pro 9n and Yoga 9n 2-in-1
Lenovo has two named RTX Spark laptops on the record: the Yoga Pro 9n and the Yoga 9n 2-in-1. Both carry the “9n” naming Lenovo has been using to signal Nvidia-powered Arm silicon inside its Yoga line, distinguishing them from the Intel and AMD variants that share the same chassis names. Neither Lenovo nor the outlets covering the announcement have published full spec sheets, pricing, or an exact on-sale date beyond the October window shared by the rest of the platform.
The Yoga 9n 2-in-1 is the more interesting of the two purely from a positioning standpoint. A convertible form factor paired with an AI chip built around “personal agents” suggests Lenovo is targeting the same crowd that buys Surface-style hybrids, rather than pitching RTX Spark strictly as a workstation-replacement play.
ASUS, Dell, HP, Microsoft Surface, MSI and GIGABYTE Join the Rollout
Beyond Lenovo and Acer, the RTX Spark partner list reads like a roll call of the entire Windows PC industry. ASUS has confirmed three RTX Spark products: the ProArt P16 and ProArt P14 laptops, plus the ProArt GR1X, a mini PC built around the chip. ASUS’s ProArt line has historically targeted creators and prosumers rather than gamers, which fits the “personal agents” and local-AI framing Nvidia is using for the platform.
Dell, HP and Microsoft Surface are all confirmed as RTX Spark PC partners, though specific model names for those three have not been published in the reporting available at press time. MSI has shown small-form-factor RTX Spark desktops alongside its laptop plans, with the company describing both as “coming this fall,” a slightly looser window than the October date attached to Lenovo and Acer. GIGABYTE is also attached to the platform, with its models described as “to follow” rather than launching alongside the first wave.
That spread across nearly every major Windows OEM is unusual for a first-generation Nvidia platform. It suggests Nvidia negotiated broad industry buy-in before making any public announcement, likely because none of these companies wanted to be the one PC maker without an answer to a chip Nvidia was already selling to developers through DGX Spark.
RTX Spark Platform Specs: What Nvidia Has Confirmed
Stripped down to what Nvidia has actually put in writing, the RTX Spark platform spec sheet is short but pointed: 1 petaflop of AI performance, up to 128GB of unified memory, and what the company calls “full-stack NVIDIA AI and graphics technology.” Nvidia has not published clock speeds, core counts, or a specific process node for the RTX Spark silicon in the material available so far, and outlets covering the announcement have generally repeated Nvidia’s own figures rather than independently benchmarking early units, since none appear to be in reviewers’ hands yet.
The three confirmed form factors are laptops, compact desktops or mini PCs, and a workstation-class client tier that Nvidia has described as part of the same RTX Spark family, built on the Spark plus Blackwell combination. That workstation tier is the piece most likely to appeal to the same buyers who might otherwise consider a DGX Spark, and it’s a category where Nvidia has given the least public detail so far.
RTX Spark vs DGX Spark: Same Silicon, Different Buyer
Nvidia’s language around RTX Spark, especially the 1 petaflop and 128GB unified memory figures, closely echoes how the company has described DGX Spark, the developer-focused desktop box it introduced earlier in 2026 as a way to give AI researchers Grace Blackwell compute without needing a data-center allocation. shattered.io covered how that same DGX Spark hardware is now being pooled alongside RTX and Mac M4-and-later hardware through Nvidia’s PAIR beta, underlining that DGX Spark is already an established, shipping product rather than a hypothetical.
The practical difference is audience and packaging. DGX Spark ships as a dedicated appliance aimed at developers who want a local AI sandbox. RTX Spark, by contrast, is built into ordinary-looking laptops and desktops running full Windows, sold through the same retail and OEM channels as any other PC. If the underlying compute really is comparable between the two, RTX Spark effectively makes DGX Spark-class AI hardware available to anyone buying a Lenovo Yoga rather than only to developers who specifically sought out an Nvidia dev box.
RTX Spark PC Launch Roster at a Glance
The table below summarizes what’s confirmed about each OEM’s RTX Spark plans as of this reporting. Status reflects the most specific timing each outlet has published; where a model name hasn’t been disclosed, that’s noted rather than guessed.
| OEM | Confirmed Models | Form Factor | Launch Status |
|---|---|---|---|
| Lenovo | Yoga Pro 9n, Yoga 9n 2-in-1 | Laptop, 2-in-1 | October (first wave) |
| Acer | Not yet named | Laptop | October (first wave) |
| ASUS | ProArt P16, ProArt P14, ProArt GR1X | Laptop, mini PC | Fall 2026 |
| Dell | Not yet named | Unconfirmed | Fall 2026 |
| HP | Not yet named | Unconfirmed | Fall 2026 |
| Microsoft Surface | Not yet named | Unconfirmed | Fall 2026 |
| MSI | Not yet named | Laptop, small-form-factor desktop | “Coming this fall” |
| GIGABYTE | Not yet named | Unconfirmed | “To follow” |
How RTX Spark Stacks Up Against Apple Silicon and AMD Ryzen AI Max
RTX Spark isn’t launching into an empty category. AMD has been pushing its own local-AI memory story with the Ryzen AI Max PRO 400, which shattered.io previously reported ships with up to 192GB of memory, enough to run large open-weight models with roughly 300 billion parameters directly on a laptop or mobile workstation. That’s a larger memory ceiling than the 128GB Nvidia has confirmed for RTX Spark, though memory capacity alone doesn’t determine which platform runs a given model faster, since architecture, bandwidth and software stack maturity all factor in.
Apple’s answer sits at the high end of the professional workstation market rather than the laptop segment. The Mac Studio built around the M5 Ultra chip, which shattered.io covered when Apple launched its M6 and M5 Ultra lineup, offers up to 512GB of unified memory and an 80-core GPU, dwarfing both RTX Spark’s and Ryzen AI Max PRO 400’s memory ceilings. Apple’s platform is priced and positioned as a desktop workstation rather than a consumer laptop chip, though, so it isn’t a direct like-for-like competitor to RTX Spark’s laptop-first push.
| Platform | Maker | Max Unified Memory | Confirmed AI Performance | Primary Form Factor |
|---|---|---|---|---|
| RTX Spark | Nvidia | 128GB | 1 petaflop (Nvidia figure) | Windows laptop / mini PC |
| DGX Spark | Nvidia | 128GB | 1 petaflop (Nvidia figure) | Developer desktop appliance |
| Ryzen AI Max PRO 400 | AMD | 192GB | Not directly comparable metric | Laptop / mobile workstation |
| M5 Ultra (Mac Studio) | Apple | 512GB | Not directly comparable metric | Desktop workstation |
Reading this table, the headline isn’t that Nvidia has the biggest number. It doesn’t. The story is that Nvidia is the only company on this list bringing its architecture to mainstream Windows laptops from Lenovo, Acer, ASUS, Dell and HP simultaneously, rather than shipping through a single branded product line the way Apple does with Mac or AMD does through its own reference designs.
Market Impact: Nvidia’s Push to Own the Whole Machine
For Nvidia, RTX Spark represents a shift in how the company captures value from a PC sale. Historically, Nvidia sold a GPU to Dell or Lenovo, and those companies built, marketed and priced the finished machine. With RTX Spark, Nvidia’s chip is the entire compute platform, CPU and GPU together, which gives Nvidia a bigger share of the bill of materials and, potentially, more control over how “AI PC” gets defined for Windows going forward, at a moment when Qualcomm, Intel and AMD are all making competing claims about their own AI PC chips.
For the OEMs, the calculation is different. Lenovo, Acer, ASUS, Dell, HP, Microsoft, MSI and GIGABYTE all get a new SKU tier they can market as agentic-AI-ready without having to develop their own silicon, something none of them, aside from Microsoft with Surface’s various custom efforts, has managed at scale. The risk for those OEMs is differentiation: if the underlying RTX Spark chip is identical across brands, the competitive battle shifts entirely to chassis design, display quality, battery life and price, the same fight that already plays out every year in the ordinary laptop market.
The Memory Shortage Shadow Over the October Launch
RTX Spark’s entire pitch rests on generous unified memory, which makes the timing awkward. shattered.io has reported that Nvidia already raised AI server prices by more than 15% this year, citing a memory supply crunch, and separately covered how RTX 5090 prices have roughly doubled as AI demand competes with gaming demand for the same memory supply. A platform built around packing 128GB of unified memory into a consumer laptop is launching directly into that shortage.
That tension could show up in RTX Spark pricing once OEMs disclose it, or in constrained initial supply if Lenovo and Acer can’t secure enough memory to meet October demand. Neither company has published pricing or unit availability targets in the reporting available so far, so it isn’t possible to say with confidence how the shortage will land on shelf prices. It is, at minimum, a variable worth watching once the first units actually ship.
Historical Context: Nvidia’s Long Road From Chips to Computers
Nvidia has flirted with building complete systems before without ever fully committing to consumer retail. The Shield line of Android set-top boxes and tablets, the short-lived reference-design push around Nvidia-branded tablets in the early 2010s, and the DGX line of AI appliances for enterprise buyers were all steps toward controlling more of the finished product. None of those, though, put Nvidia’s own branded silicon at the center of a mainstream Windows laptop sold by five or more of the industry’s largest PC makers at once.
That’s what makes RTX Spark different from those earlier efforts. Huang’s framing at Computex, that Nvidia and Microsoft are set to “reinvent the PC,” only makes sense as a claim if the platform actually reaches ordinary consumers through ordinary retail channels, not just developers buying a DGX box directly from Nvidia. The October launch is the test of whether that ambition survives contact with the retail supply chain, OEM pricing decisions and, per the previous section, a tight memory market.
What Nvidia and Industry Voices Are Saying
Nvidia’s own public statements have been the most detailed source of information on RTX Spark so far. The company’s Computex 2026 announcement described the initial rollout in broad terms: “RTX Spark laptops and compact desktops will be available this fall from leading manufacturers including ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI, with models from Acer and GIGABYTE to follow,” Nvidia said, per its own Computex 2026 announcement.
That’s worth sitting with, because it’s slightly at odds with the more recent Reuters reporting naming Lenovo and Acer as the two OEMs launching first in October. Nvidia’s own Computex-era list put Acer in the “to follow” group behind ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI. By the time Reuters filed its October-launch report, Acer had apparently moved up to join Lenovo in the first wave, while the OEMs Nvidia originally listed as launching first, ASUS, Dell, HP, Microsoft Surface and MSI, are now described in the more recent reporting as following later in the fall. Plans clearly shifted somewhere between Computex and the confirmed October announcement, and neither Nvidia nor the OEMs involved have publicly explained why the sequencing changed.
Huang’s own remarks add the ambition layer to those shifting logistics. “This is going to be the new PC,” he told the Computex audience, per Fortune’s reporting, before adding that “Microsoft and Nvidia are going to reinvent the PC.” Coming from the CEO of the company supplying the GPU inside the large majority of gaming and workstation PCs sold today, that’s a statement about wanting a bigger piece of the finished product, not just the chip inside it.
Predictions: What Comes After the October Launch
- Expect Lenovo and Acer’s first RTX Spark units to arrive with limited regional availability rather than a simultaneous global launch, consistent with how StorageReview and Reuters have both framed the rollout as staggered.
- Pricing will likely land at a premium versus comparable non-RTX-Spark laptops in the same Lenovo and Acer lines, given the cost of packing 128GB of unified memory into a laptop chassis during a memory shortage.
- ASUS’s ProArt-branded RTX Spark laptops and the GR1X mini PC will probably target creative professionals and AI developers specifically, based on ProArt’s existing market positioning, rather than competing head-on with gaming laptops.
- Watch for Dell, HP and Microsoft Surface to disclose specific RTX Spark model names before the end of 2026, closing the gap between “confirmed partner” and “named product” that currently separates them from Lenovo, Acer and ASUS.
- AMD and Apple are unlikely to sit still. Given AMD’s existing Ryzen AI Max PRO 400 memory advantage and Apple’s Mac Studio lead, both companies have incentive to push their own local-AI messaging harder once RTX Spark units are actually in reviewers’ hands and benchmarks start circulating.
Frequently Asked Questions
When exactly will RTX Spark PCs launch?
Nvidia and its OEM partners have confirmed October 2026 as the launch month for the first RTX Spark Windows PCs, according to Reuters and StorageReview. Neither Nvidia nor its partners have published an exact calendar date within October.
Which companies are making RTX Spark PCs?
Eight OEMs are currently attached to the RTX Spark platform: Lenovo, Acer, ASUS, Dell, HP, Microsoft Surface, MSI and GIGABYTE. Lenovo and Acer are confirmed to launch first, in October, with the others following later in the fall.
What are the confirmed specs of the RTX Spark chip?
Nvidia has confirmed 1 petaflop of AI performance and up to 128GB of unified memory for the RTX Spark platform. The company has not published clock speeds, core counts, or process node details in the material available so far.
Is RTX Spark the same chip as DGX Spark?
Nvidia has described both platforms using nearly identical performance figures, 1 petaflop of AI performance and up to 128GB of unified memory, suggesting a shared underlying architecture built on Grace Blackwell. DGX Spark ships as a dedicated developer appliance, while RTX Spark is built into consumer laptops and desktops running Windows.
Which named RTX Spark laptop models exist so far?
Lenovo’s Yoga Pro 9n and Yoga 9n 2-in-1 are confirmed by name. ASUS has named the ProArt P16 and ProArt P14 laptops plus the ProArt GR1X mini PC. Acer, Dell, HP, Microsoft Surface, MSI and GIGABYTE have not yet disclosed specific model names in available reporting.
How does RTX Spark compare to AMD’s Ryzen AI Max PRO 400?
AMD’s Ryzen AI Max PRO 400 offers up to 192GB of memory, more than the 128GB Nvidia has confirmed for RTX Spark. Memory capacity alone doesn’t determine real-world AI performance, since bandwidth, architecture and software support all play a role, and neither platform has published directly comparable third-party benchmarks yet.
Will RTX Spark PCs be expensive given the current memory shortage?
Nvidia and its OEM partners haven’t published RTX Spark pricing yet. Given that Nvidia has already raised AI server prices over a memory supply crunch and RTX 5090 prices have climbed for the same reason, a memory-heavy consumer platform like RTX Spark launching into that market is a real cost pressure worth watching once pricing is announced.
Is RTX Spark meant to replace a gaming laptop?
Based on Nvidia’s own framing, no. The company has positioned RTX Spark specifically around “personal agents” and local AI workloads rather than gaming performance, and the confirmed OEM models so far, including ASUS’s ProArt line, lean toward creator and productivity positioning rather than gaming branding.
