MIPS actual encoding scheme

Question

How does MIPS actually store opcodes against instructions? Does it linearly search or use hashtable?

I want to know how does it do either in digital logic terms or programming terms.

What i mean is how does "add" become 0x10. Does it perform operation on the ascii code of "add" or it has it already saved and if saved then how?

If not for MIPS then any other architecture would work.

What i have found till now is it that it forms a symbol table for these opcodes. Any links for this would be helpful

Thanks in advance.

Answer 1

As @Michael noted in his comment, the processor never sees your mnemonic (such as add or lw). In fact, some of the most common command mnemonics (such as bge, branch if greater than or equal) are actually pseudo-instructions, which the compiler converts into multiple native instructions.

If you Google "MIPS Green Sheet", you'll find the reference card that is included in the typical Computer Architecture textbook, explaining how each instruction line - mnemonic plus parameters - is converted to a 32-bit word.

If you spend some time plugging MIPS instructions into the MIPS Instruction Converter, you'll see how something like add $t0, $t1, $t2 becomes a 32-bit sequence of 1's and 0's (represented in hex as 0x012A4020), which is all the CPU ever really sees.