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

Re: [802.3_100GNGOPTX] copper and optics differences



Chris, Jeff-

Just to expand on the below:

The top 3 EDA companies are 10B of market cap and rely on high-volume, stable processes to allow the creation of stable compact model sets.  If the entire photonic design automation industry is 1/100 the size, I'd be shocked.  And of course we don't have stable high volume processes for very many things.

Simulating photonics is intrinsically easier, from a computational standpoint, than electronics, because photonics tends to be linear to first order, whereas transistors tend to be massively nonlinear.  Once we're at high enough volumes to justify the effort, I don't think we are going to have trouble creating simulation environments that are just as good as what is done in EDA for photonic chip-scale systems.

Best

M



Michael Hochberg
Director, OpSIS

On Oct 24, 2012, at 5:52 AM, Jeffery Maki <jmaki@xxxxxxxxxxx> wrote:

Chris,

I agree but something seems confused:
Photons are bosons and can occupy the same quantum state.  Electrons are fermions and cannot occupy the same quantum state.  I think electrons would thus be more like cats as they tend to fly apart the more you push them together into the same place at the same time.

The precision to which optical theory can match experiment can be quite pleasing if not astounding.  The problem is capturing all the relevant physical phenomena.  It is called "design to process" for a component that provides the challenge.  It is quite intensive to develop a model of what a given micro-fabrication facility produces that captures everything such as stress induced birefringence, density variations, scattering from imperfections in lithography, the list goes on.  Success with this approach has been established.  The direct evidence is sub-wavelength optical lithography itself for electronics.  Still, I do get the point you make.

What I do believe we have an opportunity to do is to develop an impairment model to compare the robustness of signal propagation over the fiber media including patch panels.  This could potentially be done using RSoft OptSim, VPI, or other tool with the appropriate model being setup.  Then we might have a means for objectively comparing reception and the resulting BER of competing proposals for signaling.

Jeff


-----Original Message-----
From: Chris Cole [mailto:chris.cole@FINISAR.COM]
Sent: Tuesday, October 23, 2012 10:41 AM
To: STDS-802-3-100GNGOPTX@xxxxxxxxxxxxxxxxx
Subject: [802.3_100GNGOPTX] copper and optics differences

In the Next Generation Optical Ethernet Study Group, a number of
presentations heralded the arrival of optical simulation tools that will
rival the modeling accuracy of IC simulation tools, leading to first
pass success of complex optical chips. Those of us that have been
through multiple iteration cycles designing even simple optics have
remained skeptical that our jobs are soon going to be a lot easier. This
has been reaffirmed by presented measurements.

A recent quote from Mehdi Asghari, CTO of Kotura, wonderfully captures
the difference between designing electronics and optics.

"Photons and electrons are like cats and dogs. Electrons are dogs: they
behave, they stick by you, they are loyal, they do exactly as you tell
them, whereas cats are their own animals and they do what they like. And
that is what photons are like."

If you find the cat herding analogy too simplistic an explanation, I am
confident Mehdi can expand using quantum statistical mechanics.

Chris