Performance Evaluation of Cryptographic Algorithms on Resource-constrained and GPU/TPU integrated Multi-core Processor Systems

Encryption and decryption play a crucial role in securing data in modern applications, from personal devices to enterprise-level systems. This study evaluates the performance of widely used encryption algorithms - DES, 2DES, 3DES, AES, and Blowfish - across multiple operational modes, including Cipher Block Chaining (CBC), Electronic Codebook (ECB), Cipher Feedback (CFB), Output Feedback (OFB), and Counter Mode (CTR), using various platforms with the Python language. We compare the encryption and decryption times for these algorithms using different file sizes (10KB, 100KB, 100MB, and 1GB) across six system configurations, namely on-premises gaming laptops, desktops, Raspberry PI (IoT), and cloud platforms. Our findings show that AES and Blowfish provide superior performance when encrypting large files across various cryptographic modes, while DES and its variants, despite being secure, create significant time overhead as file sizes increase. This study offers insights into how various encryption and decryption algorithms and modes perform in real-world scenarios, guiding practitioners in selecting the most efficient encryption techniques for applications that require speed, security, scalability, and sustainability. We also suggest future research directions, including optimizing encryption and decryption for IoT and cloud-centric devices.