Talk:Historical Notes on CISC and RISC
Anyone interested in RISC vs CISC should give Sun Microsystems libmath a read (it's in the BSD source trees). It's amazing how many thousands of lines of C they wrote, to do things that take a single instruction on Intel X87 CISC FPU. They also used a lot of type punning, so the code Sun wrote in the 90's might stop working with modern compilers. But X87 still works as it always has, and always will.
- (Above by Norepinephrine. Please sign posts on talk pages with 4 '~' characters.)
- There are several objections to this. A FPU is not a "computer" in the sense of RISC or CISC, it's not a central processing unit. Floating point emulation was commonly needed until some time in the 90s. (Later in embedded hardware.) The x87 was never the only FPU around; other CPU ranges had their own. (Until the Pentium, the x87 was slower than other architectures' FPUs, including SPARC's.) Finally, the big reason: If I remember right, libmath is a standard library which does a whole lot more than floating point emulation. It's also responsible for extended-precision arithmetic, for instance 100-digit numbers; a very hard thing to make high-performance. (I may be mixing it up with libm.) Eekee 18:25, 11 June 2019 (CDT)
Start a discussion about Historical Notes on CISC and RISC
Talk pages are where people discuss how to make content on OSDev.wiki the best that it can be. You can use this page to start a discussion with others about how to improve Historical Notes on CISC and RISC.