Happy to have increased your computer history knowledge!
If you mostly touch modern computers, it’s not surprising that you didn’t know, because Intel’s x87 floating-point co-processor is mostly a thing of the past. Since the i486, the floating-point has been integrated into the CPU rather than being a separate chip. You can still issue x87 floating-point instructions to the i486, or even to a modern Intel CPU, though! Nowadays, if you want floating-point stuff, you (or the compiler that compiles your program) will probably use SSE instructions instead. But I heard that some applications where the precision is very important still use x87 because it internally represents the numbers as 80 bit instead of 64 bit.
I forget; was it the 80387 or the 80487 that was just a full-blown DX (as in, FPU included on the die) version of its 80_86SX counterpart with only subtle differences in the pinout so it could disable the onboard main/SX processor socket?
Happy to have increased your computer history knowledge!
If you mostly touch modern computers, it’s not surprising that you didn’t know, because Intel’s x87 floating-point co-processor is mostly a thing of the past. Since the i486, the floating-point has been integrated into the CPU rather than being a separate chip. You can still issue x87 floating-point instructions to the i486, or even to a modern Intel CPU, though! Nowadays, if you want floating-point stuff, you (or the compiler that compiles your program) will probably use SSE instructions instead. But I heard that some applications where the precision is very important still use x87 because it internally represents the numbers as 80 bit instead of 64 bit.
I forget; was it the 80387 or the 80487 that was just a full-blown DX (as in, FPU included on the die) version of its 80_86SX counterpart with only subtle differences in the pinout so it could disable the onboard main/SX processor socket?
Edit: Found it! It was the i487 SX:
https://www.youtube.com/watch?v=cwkZz4uwcuQ