NVIDIA RTX Spark AI PC

NVIDIA RTX Spark AI PCs are a new generation of personal AI computers built for creating, AI development, gaming and local AI workflows. NVIDIA describes RTX Spark as the fusion of NVIDIA AI and RTX graphics in a single chip, designed for slim RTX laptops and small, ultra-efficient desktops. RTX Spark systems feature up to a 6,144-core Blackwell RTX GPU, up to a 20-core NVIDIA Grace CPU (Arm architecture), up to 128GB unified memory and up to 1 petaflop FP4 AI performance. CUDA runs natively on RTX Spark, allowing users to develop, prototype, fine-tune and run AI models locally.

What NVIDIA says RTX Spark can do. NVIDIA's announced capability targets include working with 3D scenes over 90GB, editing 12K video, running 120-billion-parameter language models with up to a 1-million-token context window locally, and AAA gaming at 1440p above 100 fps. Because the CPU is Arm-based, Windows runs x86 applications through Prism emulation, which Microsoft states is tuned for the microarchitecture of RTX Spark.

NVIDIA RTX Spark laptops and desktops are coming Fall 2026. Official RTX Spark laptop designs listed by NVIDIA include ASUS ProArt P16, ASUS ProArt P14, Dell XPS 16 Creator Edition, HP OmniBook Ultra 16, HP OmniBook X 14, Lenovo Yoga Pro 9n, Microsoft Surface Laptop Ultra and MSI Prestige N16 Flip AI+. Compact desktops announced so far include the ASUS ProArt GR1X Mini PC and the MSI EdgeMesa N AI+ Mini PC, with ASUS and MSI systems arriving first and Acer and GIGABYTE to follow.

Specifications, pricing and Malaysia availability are subject to official NVIDIA and brand confirmation. No official pricing has been announced for any RTX Spark system. ASUS ProArt P16 and P14 have been reported for Malaysia availability in Q3 2026.

NVIDIA RTX Spark AI PC

FAQ

Questions, answered.

What is confirmed about NVIDIA RTX Spark today, what is not, and how it differs from NVIDIA DGX Spark. Updated as NVIDIA and the brands confirm Malaysia availability.

  • What is NVIDIA RTX Spark?

    NVIDIA RTX Spark is a platform for Windows AI PCs, announced by NVIDIA and Microsoft in May 2026. It combines NVIDIA AI and RTX graphics in a single chip, the RTX Spark Superchip, designed for slim Windows laptops and compact, power-efficient desktops. Announced specifications reach up to a 6,144-core Blackwell RTX GPU, up to a 20-core NVIDIA Grace CPU (Arm architecture), up to 128GB unified memory and up to 1 petaflop of FP4 AI performance. CUDA runs natively, so developers can build, fine-tune and run AI models locally.

  • RTX Spark vs DGX Spark — what is the difference?

    These are two different NVIDIA products with similar names, and they are easy to confuse.

    NVIDIA RTX Spark is a Windows AI PC platform built on the RTX Spark Superchip, for laptops and compact desktops aimed at creators, developers and gaming.

    NVIDIA DGX Spark is a Linux desktop AI supercomputer built on the GB10 Grace Blackwell Superchip, running NVIDIA DGX OS for local AI development, fine-tuning and inference.

    Different chip, different operating system, different purpose. DGX Spark is available now from EMARQUE; RTX Spark systems are expected Fall 2026.

  • What can NVIDIA RTX Spark actually do?

    NVIDIA's announced capability targets include working with 3D scenes over 90GB, editing 12K video, running 120-billion-parameter language models with up to a 1-million-token context window locally, and AAA gaming at 1440p above 100 fps.

    These are NVIDIA's stated targets from announcement. Independent benchmarks have not been published, because NVIDIA has not yet allowed press to run their own tests on RTX Spark hardware. EMARQUE will publish real-world figures once independent testing is available.

  • Will my existing Windows applications run on RTX Spark?

    RTX Spark uses an Arm-based NVIDIA Grace CPU, so Windows runs x86 applications through Prism emulation. Microsoft states Prism has been tuned for the microarchitecture of RTX Spark.

    Applications compiled natively for Arm64 will run best, and most mainstream x86 software runs under emulation. If your work depends on specific professional, engineering or plugin-heavy software, confirm its Arm64 support before ordering. EMARQUE can check compatibility for your workflow before you commit.

  • When will NVIDIA RTX Spark be available in Malaysia?

    NVIDIA states RTX Spark systems arrive beginning Fall 2026 in select regions, with ASUS and MSI systems first and Acer and GIGABYTE following.

    For Malaysia specifically, the ASUS ProArt P16 and P14 have been reported for Q3 2026 availability by Malaysian press. ASUS has not published an official country list, and no other brand has given a Malaysia date. EMARQUE will update these pages as official Malaysia availability is confirmed.

  • How much will NVIDIA RTX Spark cost?

    No official pricing has been announced by NVIDIA or any brand, for any RTX Spark system, in any market.

    Figures circulating in the technology press are analyst estimates rather than confirmed prices, so EMARQUE does not quote them. Malaysian pricing will be published on these pages once the brands confirm it. Contact EMARQUE and we will let you know when pricing opens.

  • Which RTX Spark models have been announced?

    Laptops (8): ASUS ProArt P16 H7607, ASUS ProArt P14 H7407, Dell XPS 16 Creator Edition, HP OmniBook Ultra 16, HP OmniBook X 14, Lenovo Yoga Pro 9n, Microsoft Surface Laptop Ultra, and MSI Prestige N16 Flip AI+ (the only 2-in-1 in the lineup).

    Compact desktops: ASUS ProArt GR1X Mini PC and MSI EdgeMesa N AI+ Mini PC.

    Dell, Lenovo and HP have each confirmed an RTX Spark desktop in development but have not named them. Acer and GIGABYTE are confirmed partners with no models announced yet. NVIDIA has said 30 or more laptop designs and 10 or more desktops are in development.

  • Is RTX Spark suitable for local AI development?

    It is designed for it. CUDA runs natively on RTX Spark, so the tooling used across NVIDIA GPUs works locally, and up to 128GB of unified memory is shared between the CPU and GPU, letting larger models stay resident on the machine instead of in the cloud.

    NVIDIA positions RTX Spark for prototyping, fine-tuning and running models locally. For heavier, always-on or multi-user AI work, EMARQUE also supplies NVIDIA DGX Spark, DGX Station and NVIDIA AI servers.

  • How do I order NVIDIA RTX Spark from EMARQUE?

    RTX Spark systems have not been released yet, so these pages are informational and carry no pricing or checkout. Contact EMARQUE and we will let you know as soon as Malaysian availability and pricing are confirmed.

    EMARQUE supplies the full NVIDIA AI range in Malaysia and can advise which platform fits your workload today. WhatsApp 012 627 2280 or email business@emarque.co.