Re: [BP] how to evaluate signaling method
joe, mike,
First of all i am assuming that adaptation will be done off line,
the hspice simulation will be done on a fully optimized Rx, Tx
combination. This will mean that the simulation will have to be
done by the group which provides the Rx model.
I also assume that we will be using a modest 3-5 tap Tx which
can be simulated by a few dozen hspice primitives. Of course the
tap values will have to be provided for each case by the owner of
the Rx.
The ideas of worst casing the xtalk is to speed up the simulation.
If real data is used 10M bits is if anything too few (although too
many for hspice to do in reasonable time.) Sinusoidal interference
provides worst case interference all the time.
charles
Joe M. Abler said:
>
> Charles,
> We don't consider Hspice to be viable for simulating the complexities of high speed DFE based receivers. We don't even attempt to build the models anymore. Same is getting to be true for transmitters with latest FFE capabilities.
>
> Thanks, Joe
>
> Building on the points here and adding to my last comments, we find that we need to simulate millions of bit times (actually 10M) to get any reasonable view of xtalk effects. All the more reason Hspice becomes intractable.
>
> Thanks, Joe
>
>
> Joe Abler abler@us.ibm.com
> IBM Microelectronics Division 919-254-0573
> Technical Marketing & HSS Applications 919-254-9616 (fax)
> 3039 Cornwallis Road
> Research Triangle Park, NC 27709
Michael W. Altmann said:
> Joe, Charles,
>
> This has been our finding as well. Not to be unfair to Hspice, but the analog simulation environment uses very small timesteps (order of fs~ps) and this limits the overall time that can be simulated.
>
> Adaptive simulations will clearly need to simulate for milllions of bits.
>
> .../Mike
--
|--------------------------------------------------------------------|
| Charles Moore
| Agilent Technologies
| ASIC Products Division
| charles_moore@agilent.com
| (970) 288-4561
|--------------------------------------------------------------------|