Re: Byte ordering

Michael S wrote:

On Sat, 05 Oct 2024 18:11:55 GMT
anton@mips.complang.tuwien.ac.at (Anton Ertl) wrote:
 

Brett <ggtgp@yahoo.com> writes:

Didn’t it take a decade for the 386 to get a 32 bit OS

>
386/ix appeared in 1985, exactly 0 years after the 386.  Xenix/386 and
Windows/386 appeared in 1987.
>
- anton

 SunOS for i386 in 1988.
Netware 3x in 1990.
The later sold in very high volumes.

It deserved to do so:
For what it was doing (file/print service), it was by far the most efficient product I've even heard of!
Drew Major managed to get the total latency of the "ack network interrupt, parse incoming packet, determine that it is a read request for which the client has the required access rights, locate the requested data somewhere in the memory cache, construct a response packet and hand it off to the network card" down to 300 clock cycles.
Those clock cycles _might_ have been measured on a 486 with mostly single-cycle instructions, instead of the original 386 which needed 2+ clock cycles for lots of stuff. The point still stands, it was amazingly efficient.
Terje
--
- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"