Instruction-Level Parallelism (ILP) vs Thread-Level Parallelism (TLP)

Instruction-Level Parallelism (ILP) vs Thread-Level Parallelism (TLP)

ILP is Instruction-Level Parallelism while TLP is Thread-Level Parallelism. Both are techniques used to maximize performance of programs.
ILP is used to execute multiple program instructions in a single cycle of wide issue super-scalar processors while TLP is used to execute different threads of a program in parallel on multiprocessors.

Software developers can choose to utilize ILP to run instructions one after the other or even overlap program instructions, for instance, in Graphics processing, and they can utilize TLP to run program instructions in multiple processor threads concurrently, for instance, in databases to tolerate high amounts of I/O.

 

Instruction-Level Parallelism (ILP) vs Thread-Level Parallelism (TLP)
Wiki | thetqweb