What is difference between a simulator and emulator?
I have tried to google but I see there responses on SO that are much close to CS.
How can I understand the difference between a simulator and emulator in the context of EE? Does it means that an emulator is hardware just like a DSP kit or micro controller or FPGA kit and simulator is software like MATLAB or PSpice?
I disagree that the terminology is used inconsistently.
The key is what you are trying to achieve. A simulator tries to simulate (in electrical engineering this is usually based on our mathematical and physical understanding of underlying physics) how a certain device, code, machine, etc will behave. It uses the internal structure of a model to give us an estimate of what that model will do based on the inputs provided to the simulator\$^1\$.
An emulator doesn't model the inside (necessarily, you can do hardware emulation by simulating the hardware, with something like Cadence's Palladium series of emulators). What an emulator seeks to do is to model to the outside world what we think/know the inside of a black box will do. A processor emulator will not necessarily simulate every transistor inside to see how it behaves. It takes a known behavior and emulates it for the outside world to interact with. A software emulator might emulate all the responses of a different operating system or architecture, so you can a program compiled for OS A or architecture B on a different OS or architecture.
A CPU such as seen in hardware design will model how the outputs and inputs of a CPU/micro controller might behave, without having to simulate the insides.
\$^1\$ note that these inputs must not be inputs in the classical sense of 'current flows in here' or 'a force is applied this way'. You can also do things like eigenmode analysis in a lot of types of simulations to study the resonances of a device, without requiring an external input.
The terminology is used a bit inconsistently:
In a context of the two previous answers' explanation of what are simulators and emulators and how these are used for modeling, I'd like to add an example of the intertwined use of both.
A modeling system can include both simulating and emulating capabilities. The NS family of discreet-event network simulators was primarily conceived to be able to simulate behavior of packets in networks, both wired and wireless. Using simulators of this kind, students and researchers can study packet propagation, queues, routing etc. with existing protocols and develop new protocols.
A NS-3 simulator is special in that it most closely combines the two modeling flavors, simulation and emulation. In a simulator code, a file descriptor FdNetDevice represents a net device; it can also be used to provide emulation capabilities when it becomes programmatically associated with an underlying packet socket and receives a capability to send data on a "real" network. Another net device, a TapBridge NetDevice, allows a "real" host to participate in an ns-3 simulation as if it were one of the simulated nodes.
An ns-3 simulation can be constructed with any combination of simulated or emulated devices.
