Glioma grading plays a crucial role in determining treatment strategies and patient outcomes. Traditional machine learning models, while effective, often require extensive hyperparameter tuning and high computational resources, limiting their generalizability. In this study, we explore Hyperdimensional Computing (HDC) as a lightweight and efficient alternative for glioma grade classification. To enhance interpretability, SHapley Additive exPlanations (SHAP) analysis was applied, identifying key features that contributed to model predictions. Experimental results demonstrated that HDC achieved competitive performance while significantly reducing training time and inference latency compared to machine learning models. Moreover, HDC has shown better generalizability, requiring minimal hyperparameter tuning. The findings suggest that HDC is a promising approach for real-time and resource-efficient glioma grading.