Thin laptops are now capable of running high-performance games without the bulk and noise typically associated with gaming machines. NVIDIA showcased its next-generation PC platform, 'RTX Spark,' on October 16 at Coex in Seoul, focusing on achieving both high performance and portability.
The key lies in the integration of the graphics processing unit (GPU) and central processing unit (CPU). NVIDIA has combined its latest Blackwell RTX GPU with the Grace CPU into a single 'superchip.' This allows the two components to share up to 128GB of memory, reducing the burden of data transfer. This design enhances power efficiency, enabling the creation of thinner laptops compared to those with separate high-performance graphics cards.
An NVIDIA Korea representative stated, "Creating a thin and lightweight laptop with excellent performance has been one of our major missions. RTX Spark is a platform that transcends the limitations of existing PC platforms."
The gaming performance is also noteworthy. NVIDIA claims that high-end games can run at over 100 frames per second at 1440p resolution. The platform incorporates 'DLSS 5' technology, which uses artificial intelligence (AI) to enhance visuals and generate frames. However, actual gaming performance, heat generation, and fan noise may vary depending on the cooling design and operating environment of individual products.
AI represents another significant advancement. RTX Spark boasts up to 1 petaflop of AI computing power, allowing users to run large language models (LLMs) and AI agents directly on the laptop without needing to connect to external servers. It also supports high-resolution video editing and large 3D graphics tasks.
The NVIDIA Korea representative noted, "While PC performance has improved over the past 40 years, the user experience of executing programs with a mouse and keyboard has not changed significantly. Now, users can give goals, and AI agents will execute the necessary tasks."
The entry barrier for utilizing AI has also been lowered. Previously requiring a Linux environment and complex setups, AI agents can now be installed on Windows like regular applications. The representative explained, "The process from installation to downloading local AI models is becoming one-click."
RTX Spark laptops will be released through six global manufacturers, including Asus, Dell, HP, Lenovo, Microsoft, and MSI. Four manufacturers have already introduced products in South Korea, and Microsoft has begun pre-orders, increasing the number of participating companies to five.
There is interest in why Samsung Electronics and LG Electronics are absent from the initial launch list. An NVIDIA Korea representative stated, "It seems more about a cautious approach than a lack of technical capability. We have been in ongoing discussions with both companies and are still negotiating." However, the timeline for future product releases remains uncertain.
Domestic semiconductor companies will also participate in the memory supply chain. The NVIDIA Korea representative confirmed, "We can utilize LPDDR memory from Samsung, SK Hynix, and Micron for RTX Spark."
Widespread adoption may take time. The starting price is set at $2,599, with high-end models in South Korea reaching around 9 million won depending on memory capacity. The representative stated, "Initially, developers and creators will be the primary target," adding that the product line will expand in the future. As RTX Spark demonstrates the potential for thin gaming laptops and personal AI PCs, securing price and software compatibility will be crucial for broader acceptance.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.