24

This Wikipedia page says the following:

Most x86 processors since the Intel 80486 have had these x87 instructions implemented in the main CPU

So the above quote implies that some CPUs that were released after the Intel 80486 CPU did not have the x87 floating-point unit built in.

But what was the last CPU that did not have the x87 floating-point unit built in?

Improve this question
5
  • 3
    A few years ago intel introduced a set of embedded 32 Bit CPUs. THey where refered to as 486 alike or Pentium alike. I've been told from one of the project managers that the core is derivated from a modern x86 design, somewhat related to the many core implementation a few years ago, not a classic Pentium.
    – Raffzahn
    Commented Apr 12, 2019 at 11:25
  • 3
    @Raffzahn that was Quark, see my answer.
    – Stephen Kitt
    Commented Apr 12, 2019 at 11:31
  • 1
    It migh as well be noteworthy that new x86 implementations are not impossible, so asking for a 'last' should be accomodated by 'so far', shouldn't it?
    – Raffzahn
    Commented Apr 12, 2019 at 13:05
  • There is also the 486SX vs 486DX difference, from another Wikipedia page, where the FPU is disabled on the SX version.
    – Chris O
    Commented Apr 12, 2019 at 13:55
  • 1
    Along @Raffzahn's point, you might want to specify discrete processors. There isn't much demand for x86-compatible microcontrollers anymore but IP cores for SOCs are definitely still used. I know the LCD monitor I'm looking at has one.
    – user71659
    Commented Apr 12, 2019 at 19:19

2 Answers 2

Reset to default
61

As far as I’m aware, the last FPU-less x86-compatible CPU which could still be considered general-purpose is the Vortex86SX, released in 2007 and still available now. This is a Pentium-class CPU, capable of running any Pentium code which doesn’t require an FPU. It is targeted at embedded applications, with up to 512 MiB of RAM, and includes a PCI bus, USB, Ethernet, IDE, etc. It can run Linux.

Intel themselves produced FPU-less x86-compatible micro-controllers later still, in 2015: the Quark D1000 and D2000, 32 MHz Pentium-class MCUs with 8 KiB and 32 KiB of RAM respectively, and the Quark SE C1000, with 80 KiB of RAM. These were opportunistically targeted at IoT applications, and low-power applications in general. It is still possible to buy them, for a few more months.

The last x86-compatible desktop CPU designed without an FPU was probably NexGen’s Nx586, introduced in 1994, which was supposed to compete with the Intel Pentium but didn’t integrate an FPU initially. The last FPU-less CPU in that range was the Nx586-P133, introduced in late 1995.

Improve this answer
19

All Intel x86 CPUs since the 80486 line have included floating point instructions, i.e. everything from the Pentium* onward. So the last Intel processor to lack an on-board floating-point unit (FPU) was the 80486SX (and the embedded 80486GX).

Other manufacturers, who made 486-compatible processors, continued making non-FPU chips, aiming for the budget market. These include Cyrix's Cx486SLC, and AMD's AM486SX. A 66MHz version of the latter, the Am486SX2-66, was released in 1994, a year after Intel had released its first Pentium processor.

In order to compete with the Pentium range in the PC market, third-party manufacturers effectively had to include an on-board FPU, so there were no desktop "586" chips without floating-point instructions. Embedded devices tend to operate on a longer timescale, however.

I expect that the last manufactured x86 CPU that lacked floating-point instructions will have been an embedded chip such as the 80486GX, or the Vortex86SX mentioned in another answer.

* for further reading, see the Pentium FDIV bug

Improve this answer
4
  • The other answer links to the Wikipedia article on a 3rd party Pentium class chip that doesn't support FP
    – Martin Bonner supports Monica
    Commented Apr 12, 2019 at 8:40
  • 7
    Upvoted both these answers, because knowledge is being aware that tomatoes are a fruit, but wisdom is not putting them in a fruit salad ;)
    – Muzer
    Commented Apr 12, 2019 at 9:57
  • 1
    Technically the i486 has an FPU built in. The difference is that the i486SX chip simply had it disabled whereas the DX did not. If my memory serves correctly, the situation was that the SX was a means of not only producing a lower cost SKU, but also a way to salvage chips that had manufacturing defects in the FPU unit.
    – bjb
    Commented Apr 12, 2019 at 16:52
  • 7
    @bjb later 486SXs were fabbed with no FPU on the die at all. See this page for an interesting discussion on the topic (including the defects story).
    – Stephen Kitt
    Commented Apr 12, 2019 at 17:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .