Friday, February 19, 2010

Cisco day

Holy smokes! At noon on Wednesday Feb 17 in the Building 3 auditorium at Cisco, some 400+ attendees, 80% by remote technology, 15+ questions at the end, about evenly split between on-site attendees and those on Webex or in video conference rooms. My gracious hostess was Filomena Pereira, who had advertised the talk widely and given me great advice ahead of time.

Joel Bion, Cisco's SVP for R&D, gave me a tremendous introduction, noting that I'd been a key figure in cementing Cisco's first industrial order back in the early days (1986/7), and it accounted for two things -- in 1989 when he joined, it was about 25% of Cisco's annual revenue, and more importantly, HP provided the reference site of value for their sales efforts with other corporations. He provided this without my telling him the story -- amazing to me, but then, he's been there since the early days.... I linked it to the Medal of Defiance in terms of some of the HP exec's enthusiasm for inking this deal -- in truth, the system we built is why the Wizard award happened some years later from Smithsonian.

We had fun! A key point or two revolved around this question of what kind of 21st century corporation will prove most successful and why, and where does Cisco fit in that new world. We'll have a video of the talk online presently!


Walter Underwood said...

Chuck, you also gave a 20-something kid the responsibility for choosing a vendor. It turns out I made the right choice.

It didn't hurt that my manager, Bert Raphael, and my mentor, Tony Fanning, had both used the ARPANET at SRI.

We ordered six routers from the first production run and Cisco shipped us five. It took another couple of months before they swapped out the prototype in the Multibus experimenter cage (no covers) that connected our building to the backbone. I taped some cardboard over the power supply so that nobody would get electrocuted.

The flexibility at HP Corporate Networking was also a huge win. Dell Fischer was willing to support some crazy new incompatible company-wide network as long as it met three requirements: it was useful, it was maintainable, and there was some reasonable way to charge back enough to support it. Corporate even took over the network management six months ahead of schedule because they wanted to get more experience before they started charging.

The hard part was selling the HP Internet internally. That seems odd now, but explaining that it made everything better just didn't cut it. I ended up with a map that showed HP product lines that were geographically split, and explained that this would allow them to work together more closely.

For several years, HP had the largest private internet (the word "intranet" hadn't been invented yet). That gave us a nice start in consulting to help other people build their internets. That was handy, because the network group was all trained and ready to sell X.25 networks. Oops. It was a bit of a scramble to train up on IP networks.

Walter Underwood said...

One other bit of trivia, when we started planning the HP Internet, there were no HP computers that implemented standard TCP/IP. It would hook together the VAXen in the company. Period.

HP had networking in all their computers, but it was a not-quite-the-same, thus totally incompatible, version of TCP/IP.

Before we deployed the network, HP Labs had built a modified version of HP-UX with standard TCP/IP networking and apps. I was pretty happy about that, because I had not been looking forward to explaining to people that they needed to buy a VAX to use the network.

chuck said...

"The HP Way" always was to move the decision to the lowest possible level -- the idea being that the person closest to the problem would most likely "know" the best solution. This is a clear example where it worked extremely well!!! Great choice!

chuck said...

Change is always hard, and we were fighting several battles at once --1. Who is Cisco?
2. What is TCP/IP?
3. "We" don't believe in wide-area connectivity, aren't local area networks enough?
4. If we do networks, it should be "Token Ring" (acc. to recently arrived Dan Warmenhoven from IBM)

Walter Underwood said...

As I remember, it was a pure productivity pitch. I'm sure that you and Bert were answering the "who is Cisco?" questions afterwards.

I still remember that room, presenting to Lew Platt and a bunch of GMs. Geez, I was nervous.

The technology side was low-key, low risk. The big problem was waiting for someone to get into the business of building commercial routers instead of running Unix boxes for that.

The leap of faith was the productivity. How much does it save to get e-mail delivered in a few seconds, to get a document transferred from Fort Collins to Cupertino in a minute?

We know about that now, but how do you explain it to people who haven't seen it?

The execs decided to take the bet, and it sure did pay off.