Nvidia has introduced the core technologies and use cases for 'RTX Spark,' its next-generation premium PC platform tailored for running local AI agents. Designed primarily for AI development and content creation, the platform also stands out for enabling high-end gaming using the latest graphics technologies, including DLSS 5.
On November 6, Nvidia held a press conference for 'RTX Spark' at the JW Marriott Hotel in Seoul. The event highlighted the core technology and performance of RTX Spark, which went on sale locally in South Korea on October 16, and demonstrated real-world use cases across local AI agents, content creation, and gaming.
The presentation was delivered by Jeff Yen, Technical Marketing Director for Nvidia Asia-Pacific (APAC). Reflecting on the evolution of Nvidia's GPU architectures, CUDA, and RTX technologies over the past 25 years, Yen explained that AI agents will take center stage in the next generation of PCs.
"2026 is such an important inflection point that it can be called the year of local AI agents," said Director Jeff Yen. "AI agents are transforming the PC from a simple tool into a colleague you work alongside."
The RTX Spark platform he introduced is based on a superchip that integrates Nvidia's Blackwell RTX GPU and Grace CPU into a single module. The CPU, co-designed with MediaTek, features up to 20 cores, while the GPU and CPU share data over an NVLink-C2C interface.
This delivers up to 128GB of unified memory and 1 Petaflop (FP4) of AI compute performance. It was designed to relieve the memory capacity constraints that plagued running large AI models on conventional PCs, allowing the GPU and CPU to utilize the same memory resources across various workloads.
The primary differentiator highlighted for RTX Spark is its ability to run AI agents directly on the PC without relying on the cloud.
Director Jeff Yen cited privacy protection, reduced reliance on network connectivity, faster response times, and lower cloud API costs as key advantages of local AI. In particular, keeping both AI models and task files on the local PC allows sensitive data to be processed without transmitting it to external servers.
Scenarios requiring cloud models for complex tasks were also addressed. Nvidia's 'NeMo Switchyard' is a technology that selectively utilizes local and cloud AI models based on task complexity and performance requirements. Relatively simple tasks are processed on the PC, while cloud models step in when higher inference capabilities are needed.
Several examples of applying these technologies to real-world workflows were presented at the event. However, rather than allowing journalists to directly operate the software or measure performance, representatives demonstrated and explained the features within pre-configured environments.
First, the 'Chief of Staff' AI agent was introduced as an example of managing user schedules and work. Going beyond simple Q&A, the AI serves as a personal assistant that supports schedule verification and task management.
The 'Microduck' development agent was shown as an example supporting programming and robotics development. Nvidia explained that the advantages of a local AI environment become especially pronounced in development and training workloads that run iteratively over long periods.
For content creation, the 'Video Producer Agent' and 'Music Performance Reimagined' were demonstrated. These AI agents interface with various creative tools to assist in video production and creative processes.
"Knowing what you want to create and knowing how to actually create it are two entirely different things," Yen said, explaining that users can accomplish complex creative tasks with the help of AI agents even without mastering professional software suites.
According to Nvidia, more than 130 creative applications currently support RTX acceleration. The performance of RTX Spark can be harnessed not only in traditional creation tools like Photoshop and Premiere Pro, but also in generative AI-powered video editing and 3D content creation.
Gaming is another key area highlighted for RTX Spark. "AI developers and creators play games too," Yen emphasized, noting that RTX Spark is fully equipped for gaming.
RTX Spark supports Nvidia's core gaming technologies, including DLSS, Ray Tracing, Reflex, and G-SYNC. According to the company, 97 of the top 100 popular games at launch are playable on RTX Spark, with plans to further expand support in the future.
The flagship titles presented at the venue were 'NBA 2K27' and NCSoft's 'AION 2.'
In particular, 'NBA 2K27' was showcased as an example applying the latest DLSS 5 technology. DLSS 5's 3D-Guided Neural Rendering uses AI to enhance the expression of lighting and materials in-game. Rather than merely upscaling resolution or generating frames, it focuses on achieving more realistic lighting and detailed rendering of skin and hair.
Director Jeff Yen cited the October update for 'NBA 2K27' as an example, explaining that an optimized DLSS 5 model improved performance while maintaining visual quality. Presentation slides demonstrated frame rates increasing from approximately 65 FPS before the update to around 90 FPS after applying the new model.
'AION 2' was also presented as a game running on RTX Spark. Nvidia emphasized that its high-quality Unreal Engine 5 graphics are supported by technologies such as DLSS 4.5, allowing users to enjoy the game without needing a massive standalone gaming rig.
Beyond performance, what stood out at the event was the laptop's form factor and power efficiency. The superchip for RTX Spark laptops operates on a thermal design power (TDP) of 45–80W, and the displayed units felt noticeably lighter than conventional high-performance gaming laptops. Heating was also not significantly palpable during game demos, though this was an on-site impression rather than the result of formal temperature or power measurements.
Performance retention on battery power was also highlighted. Director Jeff Yen shared internal test results showing that the system maintains approximately 98% of plugged-in performance even when running on battery power.
These characteristics highlight how RTX Spark differs in scope from traditional gaming laptops. Rather than prioritizing pure gaming performance above all else, RTX Spark targets a premium PC platform focused primarily on AI development and content creation while still accommodating gaming.
Indeed, high pricing emerged as a central topic during the Q&A session following the presentation. While acknowledging that RTX Spark carries a high price point, Director Jeff Yen anticipated that early adopters, AI developers, and content creators would be the first to purchase the product in the current market.
"In the current market climate, no one is going to consider RTX Spark an inexpensive product," he explained. "As a premium platform, its value will depend on the tasks you intend to perform with it."
He also stated that systems combining traditional x86 PCs with discrete GPUs still enjoy ample demand, and that both product categories can coexist based on use cases. An Nvidia representative added that for users strictly focused on gaming, traditional high-performance gaming laptops might remain the better fit.
Meanwhile, RTX Spark-based devices on display included Microsoft's 'Surface Laptop Ultra,' ASUS's 'ProArt P16,' MSI's 'Prestige N16 Flip,' and Lenovo's 'Yoga Pro 9n.'
RTX Spark laptops are available through official manufacturer channels—including ASUS, MSI, Lenovo, and HP—as well as domestic online retailers in South Korea. To mark the launch, Nvidia is running a promotion offering select new buyers a 2-month subscription to Perplexity Pro along with bonus credits.
The target market for RTX Spark is clear: extending an AI ecosystem that has evolved primarily around the cloud onto personal PCs, enabling AI development, content creation, and gaming all on a single system. Going forward, proving its practical value to justify its premium price tag across a wide user base will be a key challenge for RTX Spark.
