Talk:Motorola 68881

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Computing  
WikiProject icon This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
 ???  This article has not yet received a rating on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.

To-do items include:

  • Release dates!
  • A list of all computers to use these FPUs
  • Their cost over time
  • A list of the clock rates available by year
  • Perhaps a list of some of the software packages that were the heaviest users of these FPUs, if you will, like Mathematica on the Mac II
  • And some info on software emulation
  • The first Power Macs were incompatible with 68K programs that used 68881 instructions because the emulator didn't support the FPU, and there was that Virtual 68881 product that was briefly popular. (Ironic seeing the PowerPC's strong FPU performance.) A mildly interesting footnote to the history of the 68881.

A quote from an Apple employee was that they put the FPU in the Mac II because they looked at some benchmarks and said "wow". Seemed engineer driven for sure; I'd estimate that 99% of user-hours on the Mac II family didn't see a single FPU instruction executed.

These days floating-point performance is really important for games because of all the 3D math, but games weren't a driver for the 68881. At the time, FPUs were rare, and sometimes Apple would make a Mac II computer without an FPU, aimed at consumers (like the LC and the IIsi), so the market probably wasn't large enough to take the effort. And Doom had used integer math; developers (even those few making original Mac games) probably figured integer math was good enough. Tempshill 18:52, 2 May 2005 (UTC)

Size of Registers[edit]

According to the Programmer's Reference Manual (M68000PRM.PDF @ Freescale) referenced on this page, the 6888x FPUs use "extended-precision" in all 8 FPx registers.

Extended precision is NOT the 8087's 'long double', but rather a 96-bit representation.

Commodore's Rom Kernel Reference Manual: Libraries book discusses this too in the IEEE/ffp math library chapters.

I've done the 'be bold' thing and corrected the text. Renegrade (talk) 09:01, 18 October 2011 (UTC)