Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: 10 Gigabit EMI


The operative phrase is "hitting the limit". 1.25 Gbaud is certainly not the
limit. Either is 2.125 Gbaud over 8B/10B since Seagate is shipping disk drives
with that technology... Yes with the worst case Idle pattern almost anyone is
capable of conceiving. To be honest, copper cables have significantly more
problems than fiber since there's only a single device-to-device jitter budget.
Fiber is not so bad.

Historical Fact #2: The FC Idle pattern was decided upon in 1992 or so based on
a first order requirement to maximize transition density, albeit at the expense
of EMI. Things have changed. 8B/10B CDRs are child's play, relative to SONET.

Don't fool yourself into thinking that scrambling is the be all, end all
to every interconnect in a 10 Gbps link. I agree that its a good fit with
production 10 Gbps lasers, but back on traces where the scrambled data MUST move
on parallel traces cost effectively it is far inferior to 8B/10B with the 25%

Please show me a complete proposal re: an 8B/10B-based chip-to-chip interconnect
like XAUI/XGXS Idle patterns, lane and link synchronization mechanism, deskew,
PLL requirements and noise susceptibility, clock tolerance compensation,
compromised false EOP error and undetected error rate, etc. Then we can talk.

Best Regards,

Mark Yu wrote:
> Jonathan,
> Probably the wall was built by those suboptimal "second order" choices?
> Could it be that people were able to hide those suboptimal/wrong
> "second order" choices by "convenience based, user friendly" boxes
> in the past and now we are hitting the limit of current box design?
> Either we invest and investigate how to do a super box design or it is
> high time we revisit those "second order" issues and make the right choices
> this time.
> -Mark
> =========================================================================
> Mark Yu                                         phone: 732-949-2185
> Bell Laboratories, Rm. 4E504                    fax:   732-949-9118
> 101 Crawfords Corner Rd.                        yu@xxxxxxxxxxxxxxxxxxxxxx
> Holmdel, NJ 07733, USA
> =========================================================================
> > ...
> > It is my opinion that the coding will be an important, but second order
> > effect. My opinion is that the rise times are a second order effect. My
> > opinion is that the actual bit rate is a second order effect.
> > ...
> > > In short, from an implementation perspective, we have a very fundamental
> > problem: the way we design systems today as racks of modules with arrays of
> > optical transceivers poking out through the EMI enclosure may simply not
> > hack it.
> >
> > Up to Gig E, we have had the luxury (admittedly, with a fair amount of pain)
> > of having a "convenience based, user friendly" box design. With 10 Gig, have
> > we finally hit the wall?
> >
> > jonathan
Richard Taborek Sr.                 Phone: 408-845-6102       
Chief Technology Officer             Cell: 408-832-3957
nSerial Corporation                   Fax: 408-845-6114
2500-5 Augustine Dr.        mailto:rtaborek@xxxxxxxxxxx
Santa Clara, CA 95054