The increasing connectivity of vehicles introduces security vulnerabilities. Deep learning (DL) offers promise for intrusion detection systems (IDS) in vehicle networks, but resource constraints in Electronic Control Units (ECUs) hinder their deployment. This study investigates the feasibility of using a quantized Long Short-Term Memory (LSTM) network to address these limitations. By quantizing the model, we aim to significantly reduce memory usage while maintaining acceptable detection accuracy. Our results demonstrate a significant memory reduction, with the quantized model occupying only 0.000065 MB compared to 0.004314 MB for the original model. While quantization may result in a slight performance degradation in detecting certain attack types, it offers a promising approach for deploying efficient and resource-constrained intrusion detection systems in the increasingly connected automotive landscape.