Timeline for Why are old CPUs like MOS Technology 6502 and Motorola 68000 considered better for real time systems applications than modern x86 based CPUs?
Current License: CC BY-SA 4.0
16 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Oct 27, 2020 at 6:32 | history | edited | tofro | CC BY-SA 4.0 |
added 1 character in body
|
Oct 19, 2020 at 18:51 | comment | added | supercat | @user253751: Racing the beam on the 2600 was another situation where individual cycles mattered, but the Apple Disk II is really clever in a lot of ways. IMHO, perhaps too clever in some ways (using separate 74LS595 parallel-to-serial and 74LS165 serial-to-parallel chips, rather than trying to use a 74LS299 to combine the functions, would have allowed code to use another 8 cycles every 32 to compute data to be written, eliminating the performance-robbing step of having to pre-encode data) but still amazingly powerful for its level of hardware complexity. | |
Oct 19, 2020 at 8:47 | comment | added | Stack Exchange Supports Israel | alluding* to "racing-the-beam" applications | |
Oct 18, 2020 at 18:45 | comment | added | Joshua | I encountered an old product our company had that did hard-realtime work on MS-DOS. The built the realtime scheduler into the application. | |
Oct 17, 2020 at 16:53 | comment | added | supercat | Many home computers were used as hard real-time systems when doing home-computer tasks such as writing or reading data from/to tape or (on the Apple II series) disk. Even an unexpected one-cycle delay while writing a disk sector on the Apple II would totally corrupt all the data thereon. | |
Oct 17, 2020 at 12:43 | comment | added | tofro | @MichaelGraf ... while it kills any deterministic response times for everything else. Nothing comes for free. | |
Oct 17, 2020 at 10:36 | comment | added | Michael Graf | Deactivating interrupts also allows for bit-banging I/O with well defined timing, and for deterministic response times when polling a port. | |
Oct 17, 2020 at 10:01 | history | edited | tofro | CC BY-SA 4.0 |
added 18 characters in body
|
Oct 17, 2020 at 9:23 | history | edited | tofro | CC BY-SA 4.0 |
added 11 characters in body
|
Oct 17, 2020 at 9:09 | history | edited | tofro | CC BY-SA 4.0 |
added 1340 characters in body
|
Oct 17, 2020 at 8:35 | history | edited | tofro | CC BY-SA 4.0 |
added 219 characters in body
|
Oct 17, 2020 at 8:05 | history | edited | tofro | CC BY-SA 4.0 |
added 340 characters in body
|
Oct 17, 2020 at 7:53 | history | edited | tofro | CC BY-SA 4.0 |
added 527 characters in body
|
Oct 17, 2020 at 3:57 | comment | added | Todd Wilcox | I know the 8259 best out of all the components mentioned here, but with my limited knowledge this answer rings the truest for me. Certainly I've studied the 586 architecture pretty well, and even have some experience (back in the before times) with an Amiga 3000UX with a 68030, what we might today call a rudimentary GPU, and asynchronous busses, and all I know tells me that the chipset used around a CPU has a huge impact in how the overall system works. | |
Oct 16, 2020 at 20:12 | history | edited | tofro | CC BY-SA 4.0 |
added 181 characters in body
|
Oct 16, 2020 at 19:53 | history | answered | tofro | CC BY-SA 4.0 |