Intellectual Property (IP) is a dangerously overloaded term. In the most general sense, it means any knowledge that is owned by someone. An example is a patent. The patent acknowledges the ownership of the knowledge and provides the terms of protection for it. However, in the semiconductor and EDA industries, the term is well understood to mean a design or verification unit that is pre-packed and available for licensing.
Semiconductor IP (SIP) and Design IP (DIP) are generally the same thing and often just referred to as IP, IP blocks or IP cores. It is a piece of the design, such as a processor, that has been pre-verified and can be included in someone else’s design. There is often a license fee attached to its usage and possibly a royalty for every device made that contains that piece of IP. IP has become essential for the creation of large and complex devices as it would be nearly impossible for any single company to develop all of the blocks necessary for these devices.
The most common types of IP are for processors, peripherals, interconnect and memories. They can be delivered as either soft IP blocks, which means that the models are technology independent and can be synthesized to target any implementation library or fabrication process, or hard IP, which means that the details implementation and process targeting have already been performed and the licensee cannot make any changes to it. Processors are usually provided as soft IP and analog blocks and memories are usually shipped as hard IP.
Verification IP (VIP) is a pre-package set of code used for verification. It may be a set of assertions for verifying a bus protocol, or it could be a module intended to be used within a defined verification methodology, such as UVM. This would often contain stimulus sequences, bus functional models, a set of checkers, coverage model and other things associated with a particular block in the design, such as a USB interface.