Artificial intelligence news.
Straight from the original announcements.

NVIDIA RTX Spark Windows laptops launch from $2,399 with up to 128GB memory

October 17, 2026, 23:00Source: NVIDIA
NVIDIA RTX Spark Windows laptops launch from $2,399 with up to 128GB memory
Image: NVIDIA

NVIDIA RTX Spark laptops are now available, bringing Blackwell RTX graphics, up to 128GB of unified memory and the NVIDIA AI platform to slim, premium Windows PCs. Prices start at $2,399, according to NVIDIA.

Key points
  • NVIDIA RTX Spark laptops are now available, starting at $2,399.
  • The superchip pairs a Blackwell RTX GPU with a Grace CPU, with up to 1 petaflop of AI compute and up to 128GB of unified memory.
  • ASUS, Dell, HP, Lenovo, Microsoft and MSI offer systems, with more on the way.
  • Six retail launch celebrations were held worldwide, including Berlin, where Jensen Huang appeared.
  • NVIDIA offers tested recipes and playbooks for running local AI on RTX Spark.

What RTX Spark offers

According to NVIDIA, RTX Spark brings decades of accelerated computing work into a single superchip. The chip combines an NVIDIA Blackwell RTX GPU with an NVIDIA Grace CPU. NVIDIA says it delivers up to 1 petaflop of AI compute and up to 128GB of unified memory.

The company positions the laptops as portable machines that handle local AI, content creation and gaming. Systems are available from ASUS, Dell, HP, Lenovo, Microsoft and MSI, and NVIDIA says additional systems are on the way. The starting price is $2,399.

Launch celebrations around the world

NVIDIA marked the release with events in several cities. Berlin hosted one of six RTX Spark retail launch celebrations worldwide. It took place during the Festival of Lights, an annual autumn art festival, which this year included RTX Spark-themed lighting on the Berlin TV Tower.

Fans gathered ahead of the launch at XPERION, the gaming venue inside Saturn at Alexanderplatz. Attendees tried RTX Spark laptops from PC makers and explored gaming, content-creation workflows and local AI capabilities. NVIDIA founder and CEO Jensen Huang made a surprise appearance, greeted the crowd, passed out popcorn, posed for selfies and gave away five personally signed RTX Spark laptops.

  • Beijing: a large crowd gathered at the JD MALL Shuangjing Store.
  • Istanbul: Vatan Bilgisayar Topkapı hosted interactive demos and programming led by local creators.
  • Seoul: the COEX East Lobby offered hands-on demos and evening sessions on gaming and AI.
  • San Francisco: a local Best Buy store held demos, fireside chats with local founders and an appearance by tech influencer iJustine.
  • Taipei: Muse Garden hosted the local AI community for music, discussions and product showcases.

What it means for buyers and developers

The launch gives Windows buyers a new class of thin laptops aimed at running AI workloads on the device itself rather than only in the cloud. With six manufacturers already shipping systems, buyers have a range of designs to choose from at the stated starting price of $2,399.

For developers and AI enthusiasts, NVIDIA points to NVIDIA-tested recipes, playbooks and resources for running local AI on RTX Spark. The company also directs followers to its NVIDIA Local AI newsletter for updates.

Related news

Smart City Expo 2026 draws 29,721 participants to Barcelona
BusinessNovember 6, 2026, 14:57

Smart City Expo 2026 draws 29,721 participants to Barcelona

Smart City Expo World Congress 2026 closed in Barcelona after attracting 29,721 participants from 140 countries, according to organizer Fira de Barcelona. The event…

Arm says its PC ecosystem has moved from compatibility to buyer choice
Chips and infrastructureNovember 2, 2026, 12:10

Arm says its PC ecosystem has moved from compatibility to buyer choice

Arm says the debate over Arm-based PCs has shifted from whether familiar applications will run to which Arm-powered machine people want to buy. In a blog post dated…

NVIDIA Tensor Core FP32 methods speed up Point Edge Transformer MLIP inference
Chips and infrastructureOctober 29, 2026, 16:00

NVIDIA Tensor Core FP32 methods speed up Point Edge Transformer MLIP inference

NVIDIA says Tensor Core methods that emulate FP32 matrix math can speed up machine learning interatomic potentials without retraining. In tests on the Point Edge…