John Proctor, one of Cerent’s hardware engineers, joined Cerent in July 1999, a year later than he could have come on board.
Why did John seemingly procrastinate for a year before he arrived at this upstart startup?
In the summer of 1998, John’s wife was pregnant. She wanted no part in moving to Petaluma simply for her husband to join a startup and work all manner of crazy hours. In the interim, John had been interviewed by Cerent’s hardware team and recalls, “The Fiberlane box was interesting.”
John Joins Cerent, Finally . . .
John started with Cerent in July 1999. He was hired to work for Bob Bortolotto on the video option for the Cerent 454. Four hours later, yes, in just 4 hours, John reported to Dan Jones.
During our December 2016 interview, John wryly commented that he could never keep a boss for long. But four hours has got to be some kind of record. He told me that during his seven-year career at Cisco, he had twelve different bosses. Once John started working for Massimo Pradi, he shared with his new boss that he never kept one for more than 7 to 8 months as his manager. Massimo said, “I’ll be here for 3 years, you can count on that.” But the Proctor curse came to pass: Massimo left after 6 months as John’s immediate supervisor. John told me he was called into Massimo’s office before he left for a new assignment, “You were right, John. I’m leaving.”
John’s first role at Cerent was to develop the digital filtering for the planned video card for the Cerent 454 platform. The objective for John and his colleagues was to take the incoming analog cable signal and digitize it for transmission over optical fiber. John claimed he was soon told, “There was not enough sales volume to build the product.”
But this fiction was not completely true, at least from the perspective of the sales team leader, Doug Juers. He was sent out to obtain an expected $125 million in incremental ‘454’ sales to the cable operators . And, as it turned out, even without a viable video option, Doug managed to sell almost $25 million in ‘454’ gear to the cable operators.
After the video development cancellation, John worked on the GigE card for the ‘454,’ the 10GigE card, and provided support on the troubled OC-192 line cards too.
A Career Growth Opportunity for John
Don Turner, the ‘327’ team leader, assumed the ‘454’ leadership role for Hui Liu, when Dan Jones left Cisco with his vested Cisco stock. John Proctor was given the leadership role for the ‘327’ platform, following Don’s transition, to ensure the second and third releases of the ‘327’ product came to market in support of existing customer commitments.
Tom Black was a key interface for the ‘327’ too, providing Ethernet support and guidance, as was Allen Carriker, who was based in North Carolina . Tom Hoffman and his team from the Virginia site provided the software support. Alan Repech was John’s key marketing interface for the ‘327,’ while Dave Doolan supported the product’s features introduction. Rick Knox and Joe Wackerman offered manufacturing support and John Moran provided program management to ensure the product’s development milestones were met.
A Tough Sell Into the Market
John shared what he felt he learned from the ‘327’ experience:
“The ‘327’ was the wrong form factor. Period.”
John cited some of the problems telcos had with the baby brother of the ‘454,’ including the side-to-side airflow. Telco customers preferred airflow to be from bottom to top, so as not to heat other equipment operating in adjacent equipment racks. In addition, there was not enough redundancy, as there was only one power supply instead of two. A single (non-redundant) control plane presented more indigestion for the telephone companies. In terms of practical matters, cable access to connect to the product to the outside world was poor. And from a marketing perspective, the DS1 and DS3 connections were starting to disappear in 2001 in favor of Ethernet connections.
Some of the more difficult prospects included AT&T and Cablevision, both ‘454’ users, and the tough experience with Verizon and their “whole list of complaints” about the ‘327’ product and its deficiencies as outlined above.
But many new and existing ‘454’ customers opted for the ‘327.’ John happily recalls the huge number of ‘327’ orders Cisco received from tw telecom: “Oh, yes, they were our big one.”
On cost reductions to address the higher price of the ‘327’ needed to achieve Cisco’s required margins, John said there was some cost reduction work done, but that effort was not aggressive. Maybe John’s senior management knew that platform would be replaced in the near future.
“The ‘327’ team was top notch,” John adds. “They knew they were working on a dying platform . . . but they stuck to it and instead of losing their jobs, our team secured the ‘310’ program.”
The ‘327’ was capped at Release 3, which featured the promised GigE interfaces that customers like tw telecom demanded. John estimated that Release 2 came out a year after the product’s introduction, circa January 2002.
The support engineering team, after the ‘327’ was “done,” included Jeff Ogden, David Wong, and Kevin Hayward.
For the other team members, work then commenced on the ‘310’ product, which featured a key ASIC needed to save space and achieve the 1 RU pizza box form factor. The MicroADM or MADM ASIC provided the needed savings in board real estate and power consumption. The MADM ASIC combined the BTC and SXC ASIC functionality as well as some PLL functionality.
Conflict with the Ethernet Team as the ‘310’ Came to Market
The ‘310’ team was close to hitting its price points (hence cost targets) but the Ethernet functionality drove the price/costs over the needed targets to win more business.
The Ethernet requirements demanded by Cisco compelled a $1,000 cost-based solution. Ultimately, the “Ethernet-optimized product” came to market, with this costly Ethernet option, which acted as a drag on greater sales volume.
The core team that stayed with John for the 2nd and 3rd releases of the ‘327’ were the same guys that continued on the new ‘310’ product, and included Brian Acker, Chuck Dyer, John Diab, Angelo Purugganan, Lik Seng Lim, and Hao Tran.
John’s final contribution was in October 2006. He and his Petaluma-based team worked behind the scenes to produce a working prototype for an upcoming technology bakeoff with two other competing business units within Cisco. The Petaluma engineers staged a live “Viking” demo and they blew the socks off everyone in attendance. Mike Volpi announced after the three presentations that John’s team was awarded the program.
What’s Viking you ask?
It was the product solution that eventually formed the basis for Cisco’s next world-beating product, the ASR9000.
For John, it can be said that the third time is the charm, even though he wasn’t there to taste this latest product success from Cisco.
 Doug Juers recalls his mandate to sell video and data solutions to the cable operators, “After the Cisco acquisition, we were the vertical sales force for seven quarters . . . Terry [Brown] asked me to go build a vertical account team for optical for all the cable companies. What a freakin’ challenge. I had to hire 13 people in nine months with video experience, which I hired most of my folks from Scientific Atlanta. I tried to break into all of those cable accounts in a short period of time. I had to grow the business from $400k with a quota of $25 million. I finished at $24.8M The ones we got were Cablevision, Comcast, Charter, Time Warner, AT&T Broadband (TCI), and Cox. Years later Cisco got into Adelphia with some ‘454’s after I had moved on. I only did the cable piece solely for 5 months and then I moved over under John Colvin [to work on AT&T as well].”
 Allen Carriker started with Cerent as a software engineer for Gary Baldwin’s organization in 1999. Under the Cisco umbrella, he toiled within the Optical Networking Group (ONG) as senior manager of engineering in support of the ‘454’ product development. After ONG was disbanded in 2006, Allen became a key player on the Viking (ASR9000) product development, assuming more responsibility for the program as a Senior Manager, Engineering, for another seven years. With the ASR9000 becoming part of the Service Provider Routing Group, Allen followed this organizational change to become an engineering director, in 2013, responsible for the ASR9000.