In Berlin, during the opening days of IFA 2026, Asus has unveiled the ProArt P16 and P14, the first machines in its creator-focused line built around Nvidia's RTX Spark platform. The company is betting that the next chapter of computing belongs on the desk rather than in distant data centres, offering professionals the chance to run demanding AI work on their own hardware, with their own files kept close. Yet the figures behind that promise remain Asus's own, and the question of who can afford to carry such machines, and where, is still unanswered.
Asus Bets on Local AI With Nvidia RTX Spark in New ProArt P16 and P14 Laptops
Rather than simply adding an AI badge to another laptop
Mark
Asus calls these the first ProArt laptops on Nvidia's RTX Spark. Why does that matter?
Mimi
Because the pitch is about where the work happens. Instead of sending prompts and files to the cloud, creators run AI on the laptop itself.
Mark
What does Asus say they gain?
Mimi
Three things: no token-based cloud costs, sensitive creative assets staying on the machine, and AI that still works offline.
Luke
Worth noting those are Asus's stated benefits. The announcement doesn't show any testing of them.
Mark
The headline number is 1 petaflop of AI performance. How solid is that?
Luke
It's Asus's claim, with the phrase 'up to.' And FP4 precision is part of the picture, so I wouldn't treat it as a general-purpose speed figure. No benchmarks were given.
Mimi
The memory is similar: up to 128GB unified. We don't know which configurations will actually ship.
Mark
And the physical side?
Mimi
The P16 is 12.9mm thick at 1.77kg with a 99Wh battery. The P14 is 13.9mm and 1.48kg with 90Wh.
Luke
Missing is how long those batteries last under AI loads. And there's no price or regional availability, so we can't say who can buy these.
Le Pouls
- Two new ProArt laptops arrive at IFA 2026 claiming to be the first built on Nvidia's RTX Spark platform, pressing the case for AI work done locally.
- Asus is pitching an escape from recurring cloud token costs and the exposure of sensitive creative files on remote servers.
- Headline specs of up to 1 petaflop of AI performance and 128GB of unified memory are company claims, not independently verified results.
- Creators are being offered MuseTree and StoryCube software plus agentic ComfyUI workflows to move AI tasks across generation, editing and asset management.
- Pricing and regional availability are still missing, so the real reception of these machines remains an open question.
In Berlin, during the opening days of IFA 2026, Asus has unveiled the ProArt P16 and P14, the first machines in its creator-focused line built around Nvidia's RTX Spark platform. The company is betting that the next chapter of computing belongs on the desk rather than in distant data centres, offering professionals the chance to run demanding AI work on their own hardware, with their own files kept close. Yet the figures behind that promise remain Asus's own, and the question of who can afford to carry such machines, and where, is still unanswered.
At IFA 2026 in Berlin, Asus has shown two new laptops, the ProArt P16 and the ProArt P14, which it calls the first ProArt machines built on Nvidia's RTX Spark platform. They are designed for creators and developers who would rather run AI workloads on the computer in front of them than on someone else's servers.
Inside, RTX Spark pairs a Blackwell RTX GPU with a Grace CPU and fifth-generation Tensor Cores. Asus says the platform reaches up to 1 petaflop of AI performance and 128GB of unified memory, though these are the company's own figures rather than independent measurements. The chassis stays slim: the P16 is 12.9mm thick and weighs 1.77kg, and the P14 is 13.9mm and 1.48kg. Both carry OLED panels aimed at colour-critical work, with Delta E below 1 and peak brightness of 1,600 nits.
Asus argues that local AI brings three gains: no cloud token bills, sensitive creative material that never leaves the machine, and the ability to keep working offline. Its MuseTree software runs local image models such as FLUX and WAN, StoryCube handles scene recognition and asset organisation, and a ComfyUI demonstration shows AI agents passing work between generation, editing and asset management.
What remains unknown is who will actually buy these machines. Pricing and regional availability were not announced, and until they are, the P16 and P14 stand as a serious attempt to make local AI useful to creators, with its practical reach still to be tested.
At IFA 2026 in Berlin, Asus has put two new laptops on show that it describes as the first ProArt machines built on Nvidia's RTX Spark platform: the ProArt P16 and the ProArt P14. The company is pitching them as portable workstations for creators and developers who want to run AI workloads on the computer in front of them rather than on someone else's servers.
The change that matters is inside the chassis. RTX Spark combines a Blackwell RTX GPU with 6,144 cores, an Nvidia Grace CPU with up to 20 cores, and fifth-generation Tensor Cores that support FP4 precision. Asus says the platform can reach up to 1 petaflop of AI performance and up to 128GB of unified memory. Those are the company's figures from its announcement, not independent measurements.
Asus frames the design around a different idea of what an AI PC is for. Instead of sending prompts, files and creative projects to the cloud, users would run demanding AI applications locally. The company points to three benefits: no token-based cloud costs, sensitive creative assets staying on the machine, and the ability to keep working with AI while offline.
For machines carrying that much hardware, the dimensions are modest. The ProArt P16 is 12.9mm thick and weighs 1.77kg, which Asus says makes it the thinnest 16-inch ProArt it has built. The ProArt P14 is 13.9mm thick and weighs 1.48kg. Batteries are 99Wh in the P16 and 90Wh in the P14.
The screens are aimed at professional color work. The P16 supports up to a 4K, 120Hz OLED panel with variable refresh rate, while the P14 goes up to 3K resolution. Both use what Asus calls Lumina Pro OLED technology, with a Delta E below 1 for color accuracy and peak brightness of 1,600 nits.
Asus is also bringing its own software. MuseTree integrates local image-generation models, including FLUX and WAN, so creators can generate content without paying for cloud tokens. StoryCube uses AI for scene recognition, media organisation and asset management. At the show, Asus is also demonstrating agentic workflows through ComfyUI, in which AI agents move between generation, editing and asset management instead of each tool working in isolation.
Taken together, the P16 and P14 are Asus's attempt to make local AI useful to creators rather than simply attaching an AI badge to another laptop. What the announcement leaves out is as notable as what it includes: Asus has not supplied pricing or detailed regional availability. Those details will show who these machines are actually meant for.
Citations marquantes
Asus says local AI lets creators avoid token-based cloud costs, keep sensitive assets on the machine, and work offline.— Asus, via its IFA 2026 announcement
Envie de l'histoire complète ?
Lire l'original sur Digital Trends ↗
Technology