The Intel Pentium’s instruction set is a set of instruction bytes that will always have opcode an address/register, and possibly contain a mod specifier. There will always be a primary opcode and sometimes there might be opcode used for file extensions as well. The addresses of the memory will be sent in segmented form so that there are more places that can be used with a limited number or numbers to represent where the address is. The address will consist of two four digit hexadecimal codes that will be separated by a colon, the left part being the segment and the right part being the offset. The amount of transistors that are available to use on the chip also is a big part of the instruction set, if today’s information would be put onto a chip that had only the limit of transistors from the past, it would not all fit plain and simple.

There are four groups of instruction prefixes, each with codes that are allowable for them. While the opcode is generally 1 or 2 bytes, there is the possibility if it is used for a file extension that 3 more bytes of opcode would be in the mod specifier. The specifier usually follows the primary opcode and is used as the addressing form. It contains three fields of information that may require the writer to add a SIB byte so the address of the instruction is fully specified. Segmented code may be used when referring to the memory location because it allows the machine to go to different groups and use the same numbers in each of the groups so that instead of only having one set of places that can be occupied, there is as many places available in one segment multiplied by the number of segments available for the instructions to be stored. It will take 4 bytes to hold this segmented address.

For the common values, hexadecimal form is used because it allows for more choices. Given values from 0 to 0xFF will only take up one byte instead of using other coding forms which would take up more bytes and then slow down the entire process. Depending on the machine, the amount of bytes that are used for a word can be either 2 or 4.

Since most Von Neumann architecture computers used numbers to encode there instructions, the addresses are also number and are used to go and get the instructions that are wanted to be used by the program. But you are not able to add every instruction possible to the processor since, first it is not practical since everyone wants different things, and second, it would take up too much memory and then you would slow down the program that you are trying to run. So you must have a simple instruction set that can be improved upon. In conclusion, the architecture of the Intel Pentium’s instruction set would be difficult to write is you would be trying to have all possible current instructions and were able to anticipate what the future would hold. With all the prefixes available to use, if the wrong one would be chosen then the entire process would be halted since there couldn’t be a correction until the instruction was fixed, which would hamper software production for that processor and create a really bad situation.