Terayon Founder Plans Access Revolution
Word of Rakib's latest effort started to spread at the Society of Cable Telecommunications Engineers (SCTE) Cable-Tec Expo in Atlanta last November. The company appears to lack a Web presence but Cohere does show up on this CorporationWiki page, which lists Rakib as president and Mountain View, Calif., as a company location.
Despite repeated attempts to reach Rakib in recent months, he has yet to grant an interview about Cohere.
However, several cable industry executives have heard about Cohere and have been willing to share their insights into the company's plans.
The first point to note is that Rakib appears to be in this venture without his brother Zaki, with whom he founded Terayon in 1993. It developed into a maker of cable modems, cable modem termination systems (CMTSs) and video processing and ad insertion gear. Motorola Mobility LLC bought what was left of the company (Terayon exited the Docsis business in 2005) for US$140 million in 2007, when the Rakibs were already long gone.
Post-Terayon, the Rakib brothers launched Novafora Inc., a video chip startup that flamed out in 2009. (See Motorola to Buy Terayon for $140M and Novafora Burns Out.)
Shlomo appears to be without his brother this time and is currently trying to raise funding.
As for the technology, one source identified it as "Scaling Docsis." Others note that, more specifically, Cohere's pitching an idea that would essentially gut the cable headend (where elements such as the CMTS and edge QAMs reside today) and pack those functions into small node housings and push capacity further out onto the cable network.
This more distributed approach is a radical departure from today's more centralized architecture. It would also help clear out room in cable headends that are already bursting at the seams.
There is also industry speculation that Rakib has recruited a former engineer from Infinera Corp. (Nasdaq: INFN), which pioneered the photonic integrated circuit (PIC) and knows a thing or two about how to pack a lot of technology into small spaces.
Cohere's proposal "eventually replaces most of the headend," says a cable engineer who's familiar with the project. "The more I look at it, it's quite ingenious. If it pans out, he [Rakib] might actually have something."
Fitting in with CCAP
But it's still not understood how Cohere's technology would fit with the Converged Cable Access Platform (CCAP), a super-dense architecture that combines the CMTS and edge QAM functions and puts all of cable's QAM and IP services under one roof. One engineer said the idea could remove the channelization of Docsis and make it look more like EPON Protocol Over Coax (EPoC), a budding Institute of Electrical and Electronics Engineers Inc. (IEEE) standard that aims to bring PON speeds to existing hybrid fiber/coax (HFC) networks and avoid the expense of pulling fiber all the way to the premises. (See Countdown's on for EPON Protocol Over Coax.)
Cohere is said to have a prototype to show off but is probably a couple of years away from having a commercial product.
Although Cohere's idea is getting some positive attention from cable operators, an engineer with a top-five U.S. cable operator wonders if the architecture will present significant operational challenges if it means having to touch every fiber node.
"Putting more devices on the network means more things you have to manage. It just seems very complex to me," the engineer said.
But others are already starting to buy into Rakib's vision. "He's sound in theory," says another cable engineer. "It could make your whole network digital almost overnight."
— Jeff Baumgartner, Site Editor, Light Reading Cable