The benchmarks come days before Nvidia's RTX Spark devices are expected to appear, with Microsoft's event on Wednesday, October 7 the likely venue. ASUS has separately indicated it will launch a ProArt 16 laptop with the RTX Spark N1X chip on October 8, at least in China, so the competitive window is tight. AMD's timing is no accident: it is trying to set the narrative before its rival's hardware is on the table.

The comparison AMD chose is not like-for-like. It pitted its top-spec 192GB Ryzen AI Max+ Pro 495 against Intel's Core Ultra X9 388H in a system with 64GB of memory, using ComfyUI to measure generative AI throughput. Claimed advantages range from 1.1x to 32.2x, though the 32.2x figure is a clear outlier: the model involved, Yuve, could not be found on Hugging Face, and Tom's Hardware suggests it may be an optimisation quirk or simply too large for the Intel machine.

Against the RTX Spark itself there are no published performance results, so AMD has no direct comparison to offer. The matchup so far hinges on memory: RTX Spark devices top out at 128GB of unified memory, while Gorgon Halo supports up to 192GB, letting users run larger models locally, though not necessarily faster. Tom's Hardware's own testing found the previous-gen Ryzen AI Max+ 395 underperformed the DGX Spark on both time to first token and tokens per second across three models.

Gorgon Halo is largely a refresh of Strix Halo, with the memory bump to 192GB and a 100 MHz boost clock increase on the 495 the notable changes. First devices are on sale now, with top configurations around $7,000, and AMD says it has shipped over half a million agentic PCs to date.