[sdiy] IN your mind, what is ....

[Magnus Danielson]

From: Rainer Buchty <buchty at cs.tum.edu>
Subject: Re: [sdiy] IN your mind, what is ....
Date: Sun, 1 Feb 2004 15:46:05 +0100 (CET)
Message-ID: <Pine.LNX.4.56.0402011532500.15581 at atbode100.informatik.tu-muenchen.de>

> > Actually, Zilog's Z-80 was really the same design-team as that of the
> > 8080, but doing it under their own flag, that is Zilog.
> 
> I always considered the Z80 to be the better 8080...
> 
> Add some shift registers, a little bit of memory and ROM and you get an
> entire computer including tape I/O and video output :)

Sounds like a Sinclair ZX-81 to me! ;O)

I really like the Don Lancasters advice to use a second CPU for the video-
scanning to solve the problem of the expensive CRT controllers... ;O)
Doing that would let you have a full CPU for processing. Oh happy days! ;O)

> > There is an interesting line of stuff happening around the 8080/8085
> > sucsessor [...]
> 
> That's what I meant when I said I'd like to see a parallel universe where
> "Wintel" didn't succeed. There were really interesting things going on,
> but they eventually got canned (or were later re-invented for CPU-internal
> use) with the "everything must be compatible to 1981" paradigm.

Indeed. There's been a developement in the Z-80 branch leading up to the
Rabbit CPUs.

Anyway, I think the co-processor view on things is interesting. If we play the
game again, looking back at those times and up till now. Let's look at modern
OO-problems and see what stuff we could do i hardware to make it easier.
By making it safe to do aggressive renaming and all that. I'm not sure that it
would be a total derailing. Transistors/gates are very cheap these days, so you
can do alot which you wasn't doing easilly in the early 80thies. Then the
microprocessor world started to look more and more like the minicomputer-world
with its multi-chip architectures (basically due to the need in those days
technologies, I'll probably toss that into my desk-drawer FPGA today and do it
quicker). Here's one of the fun stuff with FPGAs, you can cheaply do your own
"what if" and put it alive! Also, with the advances in compilers you can put
your own backend into a fairly complex compiler like GCC (which isn't bleeding
edge, but it is improving) without too much efforts.

> > For instance, I have one sitting in the graphical engine of my Tek 11402
> > scope. It has a 80286 for main-processor. Whoaaa! Now that's computing
> > power! ;O)
> 
> Sadly enough, any modern scope is essentially a Windows box (for the GUI)
> with some more or less sophisticated sampling unit (for measuring).

Indeed. Agilent however sells a line of logic analysers who runs a PA-RISC and
HP-UX with FVWM95 to make it "look" Windows, so there is hope! ;O)
Oh, it also works great I might add!

> > ...infact will the compacts instructions-set be good since it won't
> > trash cashes as much as 32/64 bit instructions or worse yeat - VLIW.
> 
> Aaah, that's what you have the trace cache :)

Well yes, I haven't had the time to follow all the twists! ;O)

> Not only you have the usual I/D caches with P-IV but also a trace cache
> which is used to cache already recoded (i.e. x86 to VLIW) instruction
> streams.

I haven't looked under the hood on the P-IV and I am a bit ashamed for not
doing it. OK? ;O)

> > Adding a number of other issues you really start to wonder what so great
> > about IA-64 except the amount of money spent on the project. Here the
> > Alpha stuff seems more logical. But those architectures I must confess I
> > know too little about.
> 
> It's a shame that the Alpha was essentially killed for political reasons
> after Intel took over most of the DEC team (the other part went to AMD
> which led to the K7/Athlon).

Yeap, I agree. They killed the wrong architecture!

> > The amount of money they can spend on R&D does not necessaryilly mean
> > that they will end up with a good product. It can even been argued that
> > it talks against them. It boils down to management abilities and
> > mentality of the staff to handle it well. The bigger R&D spendings you
> > have, the less accurate can each project be and it still works out one
> > way or another. It's a paradox, but there you have it, that's life.
> 
> But the more money you pump into R&D, the more "freely" research can be
> conducted and the more likely a really interesting but entirely unwanted,
> unexpected, and unplanned result may happen which leads to new insights.
> 
> If you restrict research to "improve this product" it kinda gets boring.

It comes down to management in both cases. You can do wonders with a small
budget, if you manage it well enought. You can even develop extremely good
architectures which survive individual products if you just think the right
way. If you have free development (at any budget level) you must have the
management strength to be able to "pull it in" to the development of new
products. Regardless of which budget you have, it takes a quite particular
character to be able to make good use of the resources, regardless of their
size. This particulare character is however very scarse and not often
recogniced by decission-makers. When you get the right stuff happening at the
same time, you have a potent brew thought!

> > Then some people juggle their clock to "solve" EMC problems - what a
> > broken attitude... that only works for a set of designs, and if you have
> > another design you must learn how to do it properly, so why not do it
> > properly to start with?
> 
> Sounds an awful lot like Oliver Bartels on de.sci.electronics :)

I don't know Oliver Bartels or de.sci.electronics (assuming it is in German).

A thing about EMC - many times you can reduce the problems by using fairly
simple design guidelines. By doing it early out in the design-cycle it becomes
more or less for free, and also it will become almost a second nature when
doing designs. When problems occur you are in a better situation than you would
be if you just slap in on in the "industrialization phase" of the product.

Cheers,
Magnus