What's SIMDE?

SIMDE is a simulator for supporting teaching of ILP (Instruction Level Parallelism) Architectures. It includes two characteristic flavours of ILP architectures:

• Dynamic Scheduling: Taking a Superscalar processor as an example.
• Static Scheduling: Taking a VLIW processor as an example.

The simulator places emphasis on similarities and differences between superscalar and VLIW processors because of its educative purposes.

What can't you expect from SIMDE?

SIMDE is not intended to be a realistic simulator. It contains a lot of simplifications in its design that try to clarify the functionalities at the expense of a rigorous approach. Some components are designed in such a way that makes imposible a realistic implementation.

The VLIW design is extremely simple to incide in the differences with the superscalar processor. This design is useful only as a theoretical approach.

SIMDE is not a VLIW compiler. The code optimization and scheduling are made by the user.

Functionalities

SIMDE allows to:

• Load a single sequential code each time. This code will be used as basis for superscalar executions and design of VLIW codes.

• Modify processor parameters for both Superscalar and VLIW processors.

• Create, load or modify a (very) long instruction code that can be used in VLIW simulations. Different tools (like basic blocks coloring or a code automatic-checking for detecting inconsistences) have been added in order to simplify the creation of VLIW codes.

• Modify contents from registers and memory.

• Carry out continuous or step by step simulations of Superscalar processor. The use of breakpoints is allowed to stop the execution at any point of code.

• Carry out continuous or step by step simulations of VLIW processor. The use of breakpoints is allowed to stop the execution at any point of code.