Going deep on local AI performance benchmarking with Apple’s Mac Studio with M5 Ultra taking on Nvidia DGX Spark
Apple’s latest Mac Studio with the M5 Ultra SoC has been benchmarked against Nvidia’s DGX Spark, highlighting its potential for local AI workloads.
Apple’s updated Mac Studio with the M5 Ultra SoC has already proven itself as one of the most powerful systems around, thanks to its 36-core CPU and 80-core GPU with Neural Accelerators for the matrix math operations that drive AI compute. Beyond that raw power, options for up to 512GB of unified memory have many local AI enthusiasts lining up for the 13 to 15 weeks it’ll take for Apple to ship one right now.
The system is being evaluated for its ability to handle cutting-edge large models like DeepSeek V4.1 Flash 1017 and GLM 5.3 Flash, which require significant memory resources.
Performance and value
A previous test with the Qwen 3.8 27B model did not fully exercise the 256GB of memory, prompting a new benchmark to better utilize the system's capabilities.
We wanted to fill that pool to the brim with cutting-edge large models like DeepSeek V4.1 Flash 1017 and GLM 5.3 Flash.
Market positioning
The question remains whether the tested Mac Studio configuration is worth a price tag of $12,300 or more for these users.
Related
