Key Takeaways:
- NVIDIA launched RTX Spark, a new AI-focused CPU it calls a “superchip.”
- RTX Spark delivers up to 1 petaflop of AI computing performance.
- Systems from ASUS, Dell, HP, Lenovo, Microsoft Surface, and MSI will launch this fall.
- Acer and Gigabyte models are expected later.
- NVIDIA sees a $200 billion opportunity in processors.
- The platform is designed to run AI agents and local large language models on-device.
NVIDIA AI agent PCs took center stage at Computex Taipei as the company introduced RTX Spark, a new CPU platform built to run AI agents and large language models locally. NVIDIA describes RTX Spark as a “superchip” capable of delivering up to 1 petaflop of AI computing performance, marking one of its biggest moves beyond graphics hardware.
The RTX Spark AI PC platform will debut through systems from ASUS, Dell, HP, Lenovo, Microsoft Surface, and MSI later this fall, while Acer and Gigabyte are expected to launch additional models.
A key focus of the platform is support for AI tools such as OpenClaw and Hermes Agent. NVIDIA said secure AI agent sandboxes were jointly developed with Microsoft, enabling workloads to run safely on-device. According to the company, local large language models can operate through a combination of CPU, GPU, memory, and NVIDIA CUDA software, allowing more AI tasks to be processed without relying entirely on cloud infrastructure.
NVIDIA is also targeting creators and gamers. The company says RTX technology delivers faster AI performance and AI-powered features across more than 1,000 games and applications. Support already extends to more than 100 Windows software companies, including Adobe, Blender, ComfyUI, Riot Games, and Xbox.
Following another record financial quarter, NVIDIA CEO Jensen Huang reiterated that the company sees a $200 billion market opportunity in computing processors. Huang has argued that future computing will move beyond launching applications and manually completing tasks, with users increasingly relying on AI-powered systems that respond to natural language requests.
NVIDIA’s broader strategy already includes its Vera server CPU, which the company says has generated $20 billion in sales. Huang also believes billions of AI agents will emerge over time, creating demand for more AI-capable computing devices and infrastructure.
Why It Matters:
- NVIDIA is expanding beyond GPUs into CPUs.
- AI agents are expected to drive demand for more powerful PCs.
- More AI workloads can now be processed directly on personal devices.
- The move strengthens NVIDIA’s position across PCs, software, and AI infrastructure.
NVIDIA’s move reflects growing competition to build AI PCs capable of running advanced AI workloads without depending entirely on cloud services. NVIDIA previously pursued ARM-based Windows devices, but those efforts struggled to gain traction, while Microsoft’s Surface RT resulted in a $900 million write-down.
Unlike NVIDIA’s earlier Windows ARM initiatives, RTX Spark arrives as AI workloads have become a priority for consumers, developers, and enterprises. Microsoft is branding its RTX Spark-powered device as the Surface Laptop Ultra, which it calls its most powerful Surface laptop yet.
While detailed specifications remain limited and pricing has not been widely disclosed, the systems appear closely related to NVIDIA’s DGX Spark developer platform, which currently sells for approximately $4,800.
Author’s Note: NVIDIA’s RTX Spark launch shows the company is looking beyond GPUs and toward the broader future of AI-powered computing. The move highlights why CPUs are becoming a critical battleground in the race to power next-generation AI experiences.
Leave a comment