Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Motorola Semiconductor was spun out to become Freescale. They're very much alive and kicking.

Their i.MX series of application processors are in a lot of devices.

http://en.wikipedia.org/wiki/I.MX

The old 6800/68000 cores live on in the S08/S09 and Coldfire lines.



i.MX ships are ARM architecture no?

I just think it's a shame the 68k architecture didn't last at least into the Pentium age.



Holy moly. The ColdFire. You may have had one if you had an iRiver H320 or H340 MP3 player. The Rockbox [1] guys did some amazing things with that device (and that chip).

[1]: http://rockbox.org and http://download.rockbox.org/daily/manual/rockbox-iriverh300....


I still have some ColdFires in an embedded product of the company I work in. The biggest chips even have a full-blown MMU built in. Can be made to run Linux (2.6.x.) just fine, but the toolchains (e.g. gcc) seem to have somewhat bit-rottet, unfortunately.


What's the reason that newer tool chains and kernels don't work for you?


I was using (two years ago, the last time I played with them) a freescale provided 2.6 kernel that just hasn't been upgraded, and the old mentor-graphics/code-sourcery tool-chain. Together with busybox, I was able to get a basic serial console-based system with shell, all rather straightforward.

I didn't succeed getting a newer gcc+newlib compiled for m68k, though.

All in all, the whole ecosystem around m68k seemed to have rotted away, also m68k support was also taken out of u-boot at some point if I remember correctly.

While it certainly would be possible to fix all of this, it's just too much effort "just for fun" with no particular gain besides "because I can":

If you actually have the need for some embedded CPU, one is much better off using whatever ARM core the majority of tinkerers is currently using, that way you are less likely to experience this bitrot.


There's a lot of architectures that didn't make it. It didn't just happen to Motorola.

Really, it was just a problem of a maturing market. In the early days of any industry, many bespoke options can flourish. But as time passes and the market matures, power consolidates and there is no longer room for garage-companies.


Yeah absolutely. But if you were in the early 90, the number of platforms using the 680x0 you'd think at least one of them would still be around with some very modern version of the architecture. Instead apple went PPC and the rest of the world went Intel (and finally apple too). In some ways that's what makes ARM so fun right now. Intel only is boring and everybody just send to arrive back at the PC on that architecture.


Reading the history, Motorola abandoned the 68k (Apple started switching to PPC that same year).

I don't know for sure why Moto dropped it, but lifting from Wikipedia:

While the Motorola 68040 offered the same features as the Intel 80486 and could on a clock-for-clock basis significantly outperform the Intel chip, the 486 had the ability to be clocked significantly faster without suffering from overheating problems, especially the clock-doubled i486DX2 which ran the CPU logic at twice the external bus speed, giving such equipped IBM compatible systems a significant performance lead over their Macintosh equivalents


There was a 60860 around the Pentium era, but by the mid 90s everyone had RISC fever and Moto went on with the PowerPC family.


RISC Fever started in the '80s, and Motorola's 88000 was their initial entry into the race: https://en.wikipedia.org/wiki/Motorola_88000


It wasn't just RISC fever. Motorola ran into substantial problems with getting the performance and cost to acceptable levels even for the 68060. I was an Amiga user at a time, and it was agonising for us to keep waiting for Motorola to close the performance gap that was opening to Intel CPU's. It became very clear that they were falling behind long before they announced there would not be another 680x0 CPU.


Oh you're right. For some reason I thought the line ended with the 68040.


It pretty much might have. The 68060 was way out of the price range of most people, and was in many ways a big disappointment. It delivered increased performance, but too late and at too high cost.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: