Re: [oc] Beyond Transmeta...

[Suboner@aol.com]

> great. I will take a look at it - as you probably know are simulations very
> neccessary...

I forgot to say, that under windows it depends upon GTK DLLs which can be 
found by installing Gimp or finding the DLLs amongst the developement files...

http://user.sgic.fi/~tml/gimp/win32/

> gate layer (I don't know if this is official name) = gates (like NAND) that
> can be
> executed in parallel and must wait for result from previous gate layer.

Ah, ok, that makes sense now. I sort of figured that but I was tired last 
night and couldn't think straight.

> I think there is no difference and upper statement is valid for both.

Now I am really confused :)... I was thinking about what needed to be done in 
HW because it was complicated, I guess I misunderstood what you were refering 
to and thought you ment a certain configuration/state for a network would 
need some kind of hardware for it to work, that is why I asked about it....

> :) Ok, I will rephrase my idea:
> suppose you have a NOT (gate or 1b proc.), then the following input would
> result
> outputs:
> in    out
> 0    1
> 1    0
> 1    0
> 1    0
> using my idea (2.) you get:
> in    out    'clk'
> 0    1    1
> 1    0    0
> 1    0    1
> 1    0    0
> 
> I hope this makes things clear,

Well, not really, but I know why its not clear to me, you are probably 2 or 
more steps ahead of me, because this has to do with clocks and the end of 
results, but I have not fully thought of this yet, so I cannot really 
understand what you mean, at least for right now. When I do get to that, I 
will probably look back at this and say "oh that is what he meant"... so 
anyway I'm sorry :)...

Thank you..

Leyland Needham