Performance of Programming Languages

Generations of programming languages:

1. Generation 1(A): Punched card (Binary)
2. Generation 2(B): Assembly (Text)
3. Generation 3(C): High-level (Text, eg. C, Pascal)
4. Generation 4(D): High-level with OOP (Text, eg. C++, Pascal)
5. Generation 5(E): High-level with OOP and markup (Text, eg. Qt-over-C++)

Performance of programming languages, the later the slower:

1. Assembly (top performance, but labourious development)
2. Typed languages, run natively on CPU (eg. C/C++) - Performance
1. Compiler/JIT compiler (faster)
2. Interpreter (slower)
3. Typed languages, run on VM (eg. Java) - No points
1. Compiler/JIT compiler (faster)
2. Interpreter (slower)
4. No-type languages, run natively on CPU (eg. Node.js, Cython) - Convenient
1. Compiler/JIT compiler (faster)
2. Interpreter (slower)
5. No-type languages, run on VM (eg. Python) - AI and ML
1. Compiler/JIT compiler (faster)
2. Interpreter (slower)

There's most likely no languages nowadays that has interpreter to process code files directly; interpreters compile code files to bytecode, for example: .pyc is Python compiled bytecode to be run in Python virtual machine.