The Z80A, rated for a maximum clock speed of 4 MHz, was a very widely used CPU in the 8-bit era.

The Z80H, introduced in 1986, was rated for a maximum clock speed of 8 MHz. This would at first glance appear to provide an excellent drop-in upgrade in cases where a computer is trying to be backward compatible with an older model, but now has fast enough memory chips to support the higher speed; you would get exactly double speed, which can easily be halved again for old hardware or software that depends on exact timing.

However, InfoWorld Oct 18, 1982, says, "Performance gained from the Z80H is expected to exceed that of the standard 4-MHz Z80 by four to six times. The new version of the Z80 has been designed to take advantage of 16-bit peripherals originally intended for Zilog's contender in the 16-bit race, the Z8000."

That seems to imply that it also decreases the number of clock cycles needed to carry out typical instructions. Is that the case? If it is, that would be good news for the extra performance, but bad news if you also need the ability to run at exactly the old timing; no way to have both abilities in the same chip?


The Z80H is, to my knowledge, an exact drop-in replacement for the original Z80, just with a higher clock frequency. The instruction set and the pin-out (if purchased in the same package) are 100% identical.

InfoWorld's comment would rather point into the direction that the new CPU would be able to keep up with the speed of the mentioned devices - just because of its higher clock speed. I don't see why the citation would seem to "imply a decreased number of clock cycles".

The very same article states that the improvement (to get the CPU to work at 8 MHz) was achieved by improving the process control rather than the mask in production which clearly tells me that thing is identical to a Z80A.

  • 1
    I'd guess it's confused by marketing blurb - It escapes me why doubling the clock frequency could more than double performance (maybe in constructed cases involving I/O, where you would, for example, have to wait for another revolution of a disk because you missed a sector, but definitely not in raw computing power)
    – tofro
    Dec 30 '17 at 17:15
  • 2
    Read carefully. The articles says 'performance gained from the Z80H', not 'performance of the Z80H'. Disk interleave is one obvious example where doubling the I/O speed could result in much faster system performance, particularly for a disk-bound OS such as CP/M. Dec 31 '17 at 3:00
  • 1
    @BruceAbbott Still marketing blurb - A 4 to 6 times improvement could only be achieved if you transform a badly designed I/O system based on a Z80A into an optimized system based on a Z80H.
    – tofro
    Dec 31 '17 at 13:10
  • @BruceAbbott I think that's what I'm saying - But adapting sector skewing to the I/O system's performance doesn't cost anything (except re-formatting a few disks). And it doesn't involve a new CPU.
    – tofro
    Dec 31 '17 at 13:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.