Benchmarked: Nvidia’s RTX Spark notebook delivers serious local AI compute, at a price
HP’s OmniBook Ultra 16, built around Nvidia’s N1X chip and RTX Spark platform, delivered the fastest local AI performance the review had seen in a Windows notebook. The system is aimed at people who want to run demanding AI models on their own device, and offers an alternative to Apple’s ecosystem for users who prefer Nvidia’s CUDA tools.
The tested configuration has a 20-core processor, up to 128GB of unified memory across the platform, a 2880-by-1800 OLED touchscreen and a premium aluminium chassis weighing 4.89 pounds. The review praised its performance, keyboard, trackpad and port selection, while noting that the chip runs hot and the speakers do not match Apple’s MacBook Pro. RTX Spark notebooks start at about $2,600 and can exceed $6,000, making them powerful but expensive options.
- HP’s OmniBook Ultra 16 is built for demanding local AI work.
- The review found it exceptionally fast among Windows notebooks.
- RTX Spark systems start around $2,600 and can exceed $6,000.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
For researchers, developers and other users who need to run demanding AI models locally, the OmniBook Ultra 16 offers unusually strong Windows performance and access to Nvidia’s CUDA tools. Keeping workloads on-device can also suit people who value control over their data, while the capable display, keyboard and ports add to its appeal as a premium work machine.
The case against
The high price may be difficult to justify when RTX Spark notebooks start around $2,600 and can cost more than $6,000, especially for buyers who do not need top-tier local AI performance. Heat, weaker speakers than the MacBook Pro and the notebook’s nearly five-pound weight may also make it a less compelling everyday choice than less expensive or more portable alternatives.