[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [f-cpu] asynchronous circuits
hi,
woodelf wrote:
Well this boils down to gate level logic, somthing that has to be faced
regardless of the abstract views used. From what little I have have seen
with modern CMOS logic power comsumption is used more for driving long
lines* rather than clocking, and the problem will get worse in the
future. The logic design needed is logic that is compact logic since
most of the logic designs have been developed already. Finding the non
-patented ones may be hard.
Note a long line may be 10 gates distant away since wire capactance is
geting larger than the gate capactance of the CMOS gates.
- concerning the patented designs : well, there is so much prior art for
basic things ...
- concerning long lines : that's going to plague FC0, sure. some time
ago, i thought it would be the register set....
Now, the Cell's CPUs have really large register sets and run /fast/.
And tomorrow, it is going to be something else that slows it all.
well, we would be able to measure all this if
- we had a complete source code
- we had synthesis/place/route SW...
both are very difficult to obtain, btw ;-)
YG
/away for some days again
*************************************************************
To unsubscribe, send an e-mail to majordomo@xxxxxxxx with
unsubscribe f-cpu in the body. http://f-cpu.seul.org/