It was probably hard to predict at the time, but floats larger than 64 bits never really took off. Turns out that hardware vendors didn't want to spend that man

It was probably hard to predict at the time, but floats larger than 64 bits never really took off. Turns out that hardware vendors didn't want to spend that many transistors, the x87 stack machine was a disaster, and software settled on the lowest common denominator, which was double. It's ironic that Intel was at the forefront of the floating point specification of the time, as X87 royally screwed up floating point for generations. It is was like pulling teeth to get the thing to do 64-bit math. Thank god x86 gradually moved to SSE, though it only took 25+ years to correct that fiasco.