IMPLICATOR
Nvidia set an October 2026 shipping window for RTX Spark computers. The RTX Spark lineup will have two N1X versions, with unified memory topping out at 128GB. With sales due next month, no vendor has published a price and no complete RTX Spark system or its GPU has been independently benchmarked.
What Changed
- Nvidia set an October 2026 shipping window for RTX Spark computers and used the N1X name publicly for the first time, after promising only a fall arrival at Computex in June 2026.
- The larger configuration pairs a 20-core Grace CPU with a 6,144-core Blackwell RTX GPU and 24GB to 128GB of unified memory; the smaller one uses an 18-core CPU, a 5,120-core GPU and 24GB or 32GB, and is limited to laptops at launch.
- No vendor has published a price. Lenovo says pricing for its two Yoga systems will come later, even as it gave estimated starting prices for two tablets announced the same day.
- No complete RTX Spark system or its GPU has been independently benchmarked. Reporters at IFA saw demonstrations but were not permitted to run benchmarks or change settings.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Two configurations
The larger configuration listed on September 3 combines a 20-core Grace CPU with a 6,144-core Blackwell RTX GPU and supports 24GB to 128GB of unified memory. It is intended for laptops and compact desktops. The second version pairs an 18-core CPU with a 5,120-core GPU and 24GB or 32GB of memory, and is limited to laptops at launch.
Nvidia used the N1X name publicly for the first time at IFA. At Computex in June 2026, the company had promised only a fall arrival. Eight manufacturers had systems in the lineup: Acer, Asus, Dell, Gigabyte, HP, Lenovo, Microsoft and MSI. The announced machines include Acer's compact desktop, Lenovo's Yogas and Microsoft's Surface Laptop Ultra.
The marketing block on Nvidia's product page advertises up to one petaflop of FP4 AI performance, while the specification table lists Windows 11 support. It does not list memory-channel counts, PCIe lane counts or chip model numbers.
FREE AI BRIEFING · WEEKDAYS
Follow the silicon as it reaches buyers.
Get the AI stories shaping the day, with concise context from San Francisco. The briefing takes about five minutes and arrives at 4:45 a.m. Pacific, 7:45 a.m. Eastern.
Free. No hype. Unsubscribe anytime.
Details without tests
Lenovo's September 3 announcement gives buyers a closer view of two systems. The 15.3-inch Yoga Pro 9n can be configured with up to 128GB of 9,400 MT/s LPDDR5X memory and has an 80-watt thermal limit. The 16-inch Yoga 9n 2-in-1 tops out at 64GB and 70 watts. Lenovo also lists a 92.5-watt-hour battery for the Pro and a 99.9-watt-hour battery for the convertible.
The systems had not been independently benchmarked. One reporter at IFA saw live demonstrations, but was not allowed to run benchmarks or change settings. Some display machines were switched off, and reporters were asked not to pick them up. Jon Fingas found only the core Windows and Lenovo applications on the two Yogas. "Benchmarking just wasn't an option," he wrote.
The gaming claim is 100 frames per second at 1440p with ray tracing and DLSS, in Nvidia's September 2026 material. Michael Crider noted that the figure appears beside DLSS 5, which advertises up to 6x frame generation. He "would hesitate to take that claim at face value."
Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.
Nvidia's one-petaflop AI figure in the same material uses NVFP4, a low-precision format that exchanges some accuracy for speed. Lori Grunin noted that FP4 is more prone to rounding errors than FP16 or FP32, the higher-precision formats commonly used in comparisons. Games written for x86 processors also run through Microsoft's Prism emulation layer. An Nvidia spokesperson said the performance cost should be negligible in GPU-bound games, while CPU-intensive titles could suffer larger drops.
The missing price
Lenovo's September 3 release says pricing and availability for both Yogas will come later, even as it gives estimated starting prices of 849 euros and 649 euros for two tablets announced that day. Nvidia's direct price for the unavailable 128GB DGX Spark mini-workstation was 4,800 euros as of September 2026, while Lenovo's AMD-based ThinkCentre X Ultra was expected to start at $3,699. Neither figure establishes what an N1X laptop will cost.
Josh Goldman summarized the problem: "Workstation performance in a smaller, thinner, lighter package that no one can buy 'cause the costs are too high. Nvidia has dug itself an affordability hole it can't get out of."
Frequently Asked Questions
When do Nvidia RTX Spark systems go on sale?
Nvidia set an October 2026 shipping window. It named the month rather than a specific date, and said availability will vary by manufacturer and region.
What are the two RTX Spark N1X configurations?
The larger one combines a 20-core Grace CPU with a 6,144-core Blackwell RTX GPU and supports 24GB to 128GB of unified memory, going into laptops and compact desktops. The second pairs an 18-core CPU with a 5,120-core GPU and 24GB or 32GB of memory, and is limited to laptops at launch.
How much will an RTX Spark laptop cost?
No manufacturer has published a price. Lenovo says pricing and availability for its Yoga Pro 9n and Yoga 9n 2-in-1 will be announced later. For reference, Nvidia's 128GB DGX Spark mini-workstation was priced at 4,800 euros, and Lenovo's AMD-based ThinkCentre X Ultra was expected to start at $3,699.
Has anyone tested the N1X independently?
No complete RTX Spark system or its GPU has been independently benchmarked. Reporters at IFA were shown live demonstrations but were not allowed to run benchmarks or change settings, and some display machines were switched off.
Which companies are building RTX Spark systems?
Eight manufacturers had systems in the lineup: Acer, Asus, Dell, Gigabyte, HP, Lenovo, Microsoft and MSI. Announced machines include Lenovo's Yoga Pro 9n and Yoga 9n 2-in-1, Acer's compact desktop and Microsoft's Surface Laptop Ultra.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai
