Jan 30, 2020
The History Of Cisco Welcome to the History of Computing Podcast, where we explore the history of information technology. Because understanding the past prepares us to innovate (and sometimes cope with) the future! Today we’re going to talk about the history of Cisco. They have defined the routing and switching world for decades. Practically since the beginning of the modern era. They’ve bought companies, they’ve grown and shrunk and grown again. And their story feels similar in many ways to the organizations that came out of the tail end of the grants tossed around by DARPA. These companies harnessed the incredibly innovative ideas and technology to found the companies who commercialized all of that amazing research and changed the world. These companies ushered in a globally connected network, almost instantaneously transmitting thoughts and hopes and dreams and failures and atrocities. They made money. Massive, massive truckloads of money. But they changed the world for the better. Hopefully in an irrevocable kind of way. The Cisco story is interesting because it symbolizes a time when we were moving from the beginnings of the Internet. Stanford had been involved in ARPAnet since the late 60s but Vint Cerf and Bob Kahn had been advancing TCP and IP in the 70s, establishing IPv4 in 1983. And inspired by ALOHAnet, Bob Metcaffe and the team at Xerox PARC had developed Ethernet in 74. And the computer science research community had embraced these, with the use of Email and time sharing spurring more and more computers to be connected to the Internet. Raw research being done out of curiosity and to make the world a better place. The number of devices connected to the growing network was increasing. And Stanford was right in the center of it. Silicon Valley founders just keep coming out of Stanford but this one, they were professors, and early on. They invented the multi-protocol router and finance the startup with their own personal credit cards. Leonard Bosack and Sandy K. Lerner are credited for starting Cisco, but the company rose out of projects to network computers on the Stanford campus. The project got started after Xerox PARC donated some Alto workstations and Ethernet boards they didn’t need anymore in 1980, shortly after Metcaffe left Xerox to start 3COM. And by then Cerf was off to MCI to help spur development of the backbones faster. And NSFnet came along in 1981, bringing even more teams from universities and private companies into the fold. The Director of Computer Facilities, Ralph Gorin, needed to be able to get longer network cables to get even more devices connected. He got what would amount to a switch today. The team was informal. They used a mother board from Andy Bechtolsheim, later the founder of Sun Microsystems. They borrow boards from other people. Bosack himself, who had been an ARPAnet contributor, donated a board. And amongst the most important was the software, which William Yeager wrote, which had a little routing program that connected medical center computers to the computer science department computers and could use the Parc Universal Packet (PUP), XNS, IP and CHAOSNet.. The network linked any types of computers, from Xerox Altos to mainframes using a number of protocols, including the most important for the future, IP, or the Internet Protocol. They called it the Blue Box. And given the number of computers that were at Stanford, various departments around campus started asking for them, as did other universities. There were 5,000 computers connected at Stanford by the time they were done. Seeing a potential business here, Bosack, then running the computers for the Computer Science department, and Lerner, then the Director of Computer Facilities for the Graduate School of Business, founded Cisco Systems in 1984, short for San Francisco, and used an image of the Golden Gate Bridge a their logo. You can see the same pattern unfold all over. When people from MIT built something cool, it was all good. Until someone decided to monetize it. Same with chip makers and others. By 1985, Stanford formally started a new project to link all the computers they could on the campus. Yeager gave the source to Bosack and Kirk Lougheed so they could strip out everything but the Internet Protocol and beef that up. I guess Yeager saw routers as commercially viable and he asked the university if he could sell the Blue Box. They said no. But Bosack and Lougheed were plowing ahead, using Stanford time and resources. But Bosack and Lerner hadn’t asked and they were building these routers in their home and it was basically the same thing as the Blue Box, including the software. Most of the people at Stanford thought they were crazy. They kept adding more code and logic and the devices kept getting better. By 1986, Bosack’s supervisor Les Earnest caught wind and started to investigate. He went to the dean and Bosack was given an ultimatum, it was go the wacky Cisco thing or stay at Stanford. Bosack quit to try to build Cisco into a company. Lougheed ran into something similar and quit as well. Lerner had already left but Greg Satz and Richard Troiano left as well, bringing them up to 5 people. Yeager was not one of them, even though he’d worked a lot on the software, including on nights and weekends. But everyone was learning and when it was to benefit the university, it was fine. But then when things went commercial, Stanford got the lawyers involved. Yeager looked at the code and still saw some of his in there. I’m sure the Cisco team considered that technical debt. Cisco launched the Advanced Gateway Server (AGS) router in 1986, two years after the Mac was released. The software was initially written by Yeager but improved by Bosack and Lougheed, as the operating system, later called Cisco IOS. Stanford thought about filing a criminal complaint of theft but realized it would be hard to prosecute, and ugly especially given that Stanford itself is a non-profit. They had $200,000 in contracts and couldn’t really be paying all this attention to lawsuits and not building the foundations of the emerging Internet. So instead they all agreed to license the software and the imprint of the physical boards being used (known as photomasks), to the fledgling Cisco Systems in 1987. This was crucial as now Cisco could go to market with products without the fear of law suits. Stanford got discounts on future products, $19,300 up front, and $150,000 in royalties. No one knew what Cisco would become so it was considered a fair settlement at the time. Yeager, being a mensch and all, split his 80% of the royalties between the team. He would go on to give us IMAP and Kermit, before moving to Sun Microsystems. Speaking of Sun, there was bad blood between Cisco and Stanford, which I always considered ironic given that a similar thing happened when Sun was founded in some part, using Stanford intellectual property and unused hardware back in 1982. I think the difference is trying to hide things and being effusive with the credit for code and inventions. But as sales increased, Lougheed continued to improve the code and the company hired Bill Graves to be CEO in 1987 who was replaced with John Mordridge in 1988. And the sales continued to skyrocket. Cisco went public in 1990 when they were valued at $224 million. Lerner was fired later that year and Bosack decided to join her. And as is so often the case after a company goes public, the founders who had a vision of monetizing great research, were no longer at the startup. Seeing a need for more switching, Cisco acquired a number of companies including Grand Junction and Crescendo Communications which formed like Voltron to become the Cisco Catalyst, arguably the most prolific switching line in computing. Seeing the success of Cisco and the needs of the market, a number of others started building routers and firewalls. The ocean was getting redder. John Mays had the idea to build a device that would be called the PIX in 1994 and Branley Coile in Athens, Georgia programmed it to become a PBX running on IP. We were running out of IP addresses because at the time, organizations used public IPs. But NAT was about to become a thing and RFC 1918 was being reviewed by the IETF. They brought in Johnson Wu and shipped a device that could run NAT that year, ushering in the era of the Local Area Network. John T. Chambers replaced Mordridge in 1995 and led Cisco as its CEO until 2015. Cisco quickly acquired the company and the Cisco PIX would become the standard firewall used in organizations looking to get their computers on the Internets. The PIX would sell and make Cisco all the monies until it was replaced by the Cisco ASA in 2008. In 1996, Cisco's revenues hit $5.4 billion, making it one of Silicon Valley's biggest success stories. By 1998 they were up to $6B. Their stock peaked in 2000. By the end of the dot-com bubble in the year 2000, Cisco had a more than $500 billion market capitalization. They were building an industry. The CCNA, or Cisco Certified Network Associate, and CCNE, Cisco Certified Network Engineer were the hottest certifications on the market. When I got mine it was much easier than it is today. The market started to fragment after that. Juniper came out strong in 1999 and led a host of competitors that landed in niche markets and expanded into core markets. But the ASA combined Cisco’s IPS, VPN concentration, and NAT functionality into one simpler box that actually came with a decent GUI. The GUI seemed like sacrilege at the time. And instead of sitting on top of a network operating system, it ran on Linux. At the top end they could handle 10 million connections, important once devices established and maintained so many connections to various services. And you could bolt on antivirus and other features that were becoming increasingly necessary at various layers of connectivity at the time. They went down-market for routing devices with an acquisition of Linksys in 2003. They acquired Webex in 2007 for over $3 billion dollars and that became the standard in video conferencing until a solid competitor called Zoom emerged recently. They acquired SourceFire in 2013 for $2.7B and have taken the various services offered there to develop Cisco products, such as the anti-virus to be a client-side malware scanning tool called Cisco AMP. Juniper gave away free training unlike the Cisco training that cost thousands of dollars and Alcatel-Lucent, Linksys, Palo Alto Networks, Fortinet, SonicWall, Barracuda, CheckPoint, and rising giant Huawei led to a death by a thousand competitors and Cisco’s first true layoffs by 2011. Cisco acquired OpenDNS in 2015 to establish a core part of what’s now known as Cisco Umbrella. This gives organizations insight into what’s happening on increasingly geographically distributed devices; especially mobile devices due to a close partnership with Apple. And they acquired Broadsoft in 2017 to get access to even more sellers and technology in the cloud communication space. Why? Because while they continue to pump out appliances for IP connectivity, they just probably can’t command a higher market share due to the market dynamics. Every vendor they acquire in that space will spawn two or more new serious competitors. Reaching into other spaces provides a more diverse product portfolio and gives their sellers more SKUs in the quiver to make quotas. And pushes the world forward with newer concepts, like fog computing. Today, Cisco is still based in San Jose and makes around $50 billion a year in revenue and boasts close to 75,000 employees. A lot has happened since those early days. Cisco is one of the most innovative and operationally masterful companies on the planet. Mature companies can have the occasional bumps in the road and will go through peaks and valleys. But their revenues are a reflection of their market leadership, sitting around 50 billion dollars. Yes, most of their true innovation comes from acquisitions today. However, the insights on whom to buy and how to combine technologies, and how to get teams to work well with one another. That’s a crazy level of operational efficiency. There’s a chance that the Internet explosion could have happened without Cisco effectively taking the mantle in a weird kind of way from BBN for selling and supporting routing during the storm when it came. There’s also a chance that without a supply chain of routing appliances to help connect the world that the whole thing might have tumbled down. So consider this: technological determinism. If it hadn’t of been Cisco, would someone else have stepped up to get us to the period of the dot com bubble? Maybe. And since they made so much money off the whole thing I’ve heard that Cisco doesn’t deserve our thanks for the part they played. But they do. Without their training and appliances and then intrusion prevention, we might not be where we are today. So thank you Cisco for teaching me everything I know about OSI models and layers and all that. And you know… helping the Internet become ubiquitous and all. And thank you, listener, for tuning in to yet another episode of the history of computing podcast. We are so very lucky to have you. Have a great day!