Here I offer you my top picks for what to see, smell, taste, and do at DesignCon 2017 , happening at the Santa Clara Convention Center in Santa Clara, CA, from Tuesday, January 31, through Thursday, February 2. You can register for the conference here , and see the agenda here .
1. The annual opening night #JitterPanel—a.k.a., The case of the Closing Eyes: 400G Signal Integrity —isn’t really about jitter anymore but we can’t manage to let go of the 14 year-old nickname—so what will it be this year?
We’ll start with some problems in the development PAM4 signaling from the perspective of designer and the test & measurement folks who try to satisfy them. We’ll expand to great gripes like: What’s the point in technology standards that are released a year after designs come to market? How many “goofy measurements” are crammed into specifications and which can we ignore for PAM4? How important is it to get your Gray coding right? Is FEC (forward error correction) really a cornucopia, or is it worth the cost in bandwidth, power, and latency…and BER (bit error ratio) measurement confusion? Why do we have to assume worst case patterns AND worst case crosstalk AND how many people can we cram onto that stage?
Please bring some problems (and petty grievances) for the panelists to discuss.
2. Will Ansys bring back the Maxwell’s Equations T-shirts?
Back in the good old daze, when jitter was our worst nightmare and the tech economy was still emerging from the dot-com crash, we could look forward to donning our Ansys T-shirts that proclaimed to the world, “It all began with Maxwell.” We could walk the streets in arrogance, intimidating the other kids with the vector form of the big four. But now? You can’t even make out the curl of Faraday’s law on this beat up old shirt.
I’ve heard rumors that you can get a Maxwell’s Equations shirt if you know the password…if anyone finds out the password, please post it in the comments section.
3. Is PAM4 working?
In 2015 we turned our backs on NRZ signaling—the only scheme that’s ever worked in high speed serial signaling—like Britain turned on the EU—the only signaling scheme that’s ever maintained more than two consecutive decades of peace in Europe—and rushed into a PAM4 future like (wisecrack including red baseball caps deleted).
Was the reported ease of PAM4 fake news?
Last year (2016) we saw that PAM4 is no picnic, that the mishmash of analysis techniques are silly, and that most of us have forgotten the difference between BER and SER (symbol error ratio), not to mention baud and bps.
There are twelve different presentations that have PAM4 in their titles at DesignCon2017 and at least five more that will touch on it in one way or another.
4. Will Yuri Shlepnev bring enough chocolate?
If you’ve never seen Dr. Shlepnev’s animations of electromagnetic fields and current density, then pull up a chair at the Simberian booth, grab his bag of chocolate, and let Yuri’s simulations blow you away.