ASUS Unveils ProArt P16, P14 and GR1X PCs Built Around NVIDIA RTX Spark for Local AI Workflows

On September 2, ASUS opened a private showroom at the ABION Spreebogen Waterside Hotel in Berlin, running through September 5, and used it to announce three new computers built around NVIDIA's RTX Spark chip: the ProArt P16 and P14 laptops and the ProArt GR1X mini PC. The showroom is open to invited media and partners and comes ahead of the official IFA exhibition. ASUS describes it as its first global media showcase focused on AI agents and real-world AI workflows running directly on ProArt machines.

The chip is the center of the pitch. According to ASUS, the RTX Spark superchip pairs an NVIDIA Blackwell RTX GPU with 6,144 cores and an NVIDIA Grace CPU with up to 20 cores, linked by NVIDIA's NVLink-C2C interconnect. It also includes fifth-generation Tensor Cores with FP4 precision. The company cites up to 1 petaflop of AI performance and up to 128GB of unified memory. It says that capacity lets a machine run a language model of up to 120 billion parameters, render 3D scenes larger than 90GB, and generate 4K AI video locally. These figures come from the vendors themselves, not from independent testing.

ASUS offers three arguments for running AI on the device instead of in the cloud. First, local capacity turns variable pay-per-use costs into a steady expense, and users can run and retry agents without token limits. Second, prompts, files and creative assets stay on the machine, and local models keep working offline. Third, users can match models to specific tasks, tune them through prompts or fine-tuning, and choose their own tools. These are the company's claims, not findings.

Software is part of the package. ASUS preloads MuseTree, which integrates the FLUX and WAN image models for token-free local generation, and StoryCube, which handles scene recognition, media organization and asset management on Windows. ProArt also pairs local AI with cloud services, including Adobe Creative Cloud bundles, and eligible laptops get a Google AI bundle. At IFA, ASUS says it will bring workflows validated by the industry to its own applications and to ComfyUI, an open-source generative AI workflow platform. That includes interaction between an ASUS-exclusive tool called Zenni Claw and MuseTree.

The laptops lean on the chip's integrated design for their size. The ProArt P16 measures 12.9 mm thick and weighs 1.77 kg, which ASUS says makes it the thinnest 16-inch RTX Spark laptop and the thinnest 16-inch ProArt yet, 13% thinner and 16% lighter than its predecessor. It carries a 99Wh battery. The P14 is 13.9 mm thick and weighs 1.48 kg, billed as the lightest ProArt ever, with a 90Wh battery. Both use ASUS Lumina Pro OLED displays rated at Delta E below 1 and up to 1,600 nits. The P16 offers up to 4K at 120Hz with variable refresh and NVIDIA G-SYNC, and the P14 goes up to 3K. They come in Nano Black and Neo White and include a haptic touchpad and enough ports to avoid dongles.

The GR1X is a palm-sized desktop measuring 150 × 150 × 51 mm, aimed at always-on AI work. Its dual-fan cooling uses 218 fan blades, ASUS says offers up to 1.6x the thermal coverage, and has seven fan-speed levels for sustained 24/7 running. It has 10GbE networking, WiFi 7 and Bluetooth 5.4, and supports up to four 4K displays.

The showroom also covers two other lines. The Zenbook 14 has an all-metal build with a Ceraluminum lid in Komodo Coral, Arctic Blue and Zabriskie Beige, and runs on Snapdragon X, Intel Core Ultra Series 3 or AMD Ryzen AI 300 Series chips. The ExpertBook Ultra starts at 0.99 kg and offers up to an Intel Core Ultra X9 Series 3 processor, a 14-inch 3K tandem OLED touchscreen at up to 1,400 nits, a 70Wh battery rated for up to 26 hours, and firmware security that ASUS says is PQC and NIST SP 800-193 compliant.

Visitors move through seven interactive zones and keynote sessions. Whether the performance, battery and software claims hold up will only be clear once independent reviewers get the machines.