analiza aiqbench

This benchmark evaluates the inference capabilities of the MiniMax M2.7 model when deployed on the SambaRack SN50 hardware, providing concrete data on latency and throughput for this specific AI stack. The results are relevant for developers assessing the practical performance of open-weight models on specialized inference accelerators rather than general consumer electronics or networking standards.