In an age defined by escalating cyber threats and heightened privacy concerns, Trusted Execution Environments (TEEs) have emerged as pivotal instruments for fortifying the security and confidentiality of digital domains. This paper compares commercial implementations using a conceptual model for TEEs, offering a structured framework for understanding and designing TEE-based solutions. We start by conducting a survey of existing TEE technologies, such as Intel SGX, ARM TrustZone, AMD SEV, and Apple Secure Enclave. Through a discerning comparative analysis over a conceptual reference model, we delineate the inherent strengths and limitations of each TEE, pointing their distinct features like attestation mechanisms, secure boot procedures, support for virtualization, and defenses against side channel attacks. Furthermore, our study extends to a broad spectrum of TEE applications, ranging from attestation and data protection to enabling privacy-preserving machine learning. Ultimately, this paper underscores the importance for sustained research and innovation within the domain of TEEs. Finally, we provide a discussion on these commercial implementations, methods, possible implementations for improving the security and privacy on current and future computer system applications.