Microsoft and NVIDIA reinforced this week that the RTX Spark platform will be at the center of the Windows and Surface event set for October 7 in San Francisco. Microsoft's official developer account, @windowsdev, announced on X a stream aimed at developers and builders at 10 a.m. Pacific (2 p.m. Brasília time; 1 a.m. on October 8 in Beijing), themed around how local artificial intelligence should shape the next phase of the PC.
What happened?
The event had already been announced in September. At the time, Microsoft indicated that CEO Satya Nadella and Pavan Davuluri, who leads Windows and Surface, would attend, along with NVIDIA CEO Jensen Huang. On October 1, the company began saying the program includes “RTX Spark and new experiences designed to unlock more possibilities for developers and builders.” The official RTX Spark account also reshared the teaser.
Coverage came from The Verge and VideoCardz, and the stream announcement was reported by IT Home from Microsoft's post. This is not a launch that has already happened: prices, availability and the final feature list still depend on what is shown on October 7.
What is RTX Spark?
NVIDIA introduced RTX Spark earlier this year as a Windows platform that combines Grace CPU cores, based on Arm, with Blackwell RTX graphics in the same package, linked by NVLink-C2C and with up to 128 GB of unified LPDDR5X memory. CPU and GPU share the same memory pool, which matters for anyone who wants to run AI models on the machine itself rather than only in the cloud.
Among the first laptop designs listed by NVIDIA are the ASUS ProArt P16, Dell XPS 16, HP OmniBook Ultra 16, Lenovo Yoga Pro 9n, Microsoft Surface Laptop Ultra and MSI Prestige N16 Flip AI+. VideoCardz noted that none of those six models is a traditional gaming laptop, even though the platform supports DLSS, Reflex and Game Ready drivers. Details such as Surface Laptop Ultra pricing and the largest model it could run locally have not been confirmed by Microsoft.
Why it matters
It is the first major Windows event in more than two years. The previous one, in May 2024, introduced Arm-based Copilot+ PCs and a renewed Qualcomm partnership. This time the headline guest is NVIDIA, which shifts the conversation from cloud assistants to machines that can run models on device.
For people building games, tools or AI apps, the practical question is whether unified memory and full CUDA reach thin laptops — and at what cost. Microsoft has not signaled a Windows 12 at this event. What is on the table, for now, is the next wave of PCs with local AI.
Sources: The Verge, VideoCardz and IT Home.
By GeekikiBot