An apparent engineering sample of Nvidia’s forthcoming RTX Spark processor has surfaced in a Cinebench 2026 result, offering the first glimpse of real-world performance for the Arm-based chip previously known as N1X. The entry was posted online alongside a screenshot showing the processor running under Windows 11 with the identifier JMVW0A-Generic-CPU, a label already linked to Nvidia’s Windows-on-Arm development platform.
The configuration strongly aligns with the RTX Spark specification Nvidia has outlined. The chip registers as a 20-core part and posts a multi-core score of 5,771 points, while its single-core result reaches 540 points in Maxon’s latest benchmark. Power limits reported in the leak are set at 80 W PL1 and 95 W PL2, figures that correspond precisely to the previously communicated power envelope for the RTX Spark.
Early multi-core standing among rivals
When measured against established entries in the Cinebench 2026 database, the engineering sample places competitively but not at the outright lead. It trades blows with AMD’s Ryzen AI Max+ 395 and Intel’s Core Ultra X9 388H in multi-threaded workloads. At the same time, Qualcomm’s Snapdragon X2 Elite XE2-94-100 and Apple’s M5 Max deliver notably higher numbers in both single-core and multi-core performance, underscoring the challenging Arm-based competition this generation.
System context and hardware readiness
The benchmark run was recorded on a device identified as a Surface Laptop Ultra, one of several form factors expected to carry the RTX Spark. Nvidia has confirmed the processor is destined for laptops, mini PCs, and compact AI workstations, positioning it alongside AMD’s Strix and Gorgon Halo platforms in the high-performance thin-and-light segment.
Several indicators caution against reading these results as final. The tool reports a frequency of just 1.01 GHz, an implausible readout that points to immature firmware. Both the CPU and GPU identifiers remain generic, reinforcing that this is a pre-release platform still undergoing optimisation.
Maturation potential and competitive outlook
With further firmware and driver refinements, the final silicon could narrow the gap to Intel’s Panther Lake and possibly AMD’s Medusa Point designs. Still, the current Arm heavyweight offerings from Qualcomm and Apple set a high bar that appears difficult to surpass in this cycle. Engineering sample numbers provide only a provisional baseline, but they confirm Nvidia is executing on its plan to bring a credible high-core-count Arm processor to Windows laptops and compact desktops.
Source: x.com