Micro Center’s comparison found that some larger local AI models ran on the 128GB Surface Laptop Ultra but exceeded the 32GB unit’s memory limits. Its Cyberpunk 2077 test also recorded higher base frame rates on the 128GB configuration, though the two systems have different CPU and GPU core counts.
The two configurations and their reported results
| Dimension | 32GB configuration | 128GB configuration |
| CPU | 18 cores | 20 cores |
| GPU | 5,120 CUDA cores | 6,144 CUDA cores |
| Local AI models | Some larger models exceeded this unit’s memory limits | Some larger models ran on this unit |
| Cyberpunk 2077 base frame rate | Around 37 to 38 fps | Around 46 to 47 fps |
The 32GB and 128GB figures describe shared memory, the pool available to the processor and graphics hardware. In Micro Center’s tested configurations, the extra capacity let some larger local models run where the 32GB unit hit its memory limit. That difference matters if a model you plan to use does not fit within the smaller configuration’s memory.
What the gaming results say
Micro Center recorded a higher Cyberpunk 2077 base frame rate on its 128GB unit, but memory capacity is not the only hardware difference: that configuration also has a 20-core CPU and 6,144 CUDA cores, compared with 18 CPU cores and 5,120 CUDA cores in the 32GB unit. The results therefore do not isolate memory as the cause of the frame-rate gap, and they describe this game test rather than gaming performance across all titles.
If your local-AI work involves models that exceeded the 32GB unit’s limits in Micro Center’s comparison, the 128GB configuration is the one that fits those reported tests. The Cyberpunk 2077 figures alone do not establish a universal gaming winner. For broader product context, see NeoTeo’s Surface Laptop Ultra coverage of local AI and gaming.
