Last Updated on September 16, 2023 by Dave Farquhar
3Com was a high flying maker of network cards and other network equipment. And then they disappeared with a whimper. What happened to 3Com?
3Com at the dotcom boom
Around the time of the dotcom boom, 3Com was a 5.8 billion dollar company. They were founded in 1979 by Bob Metcalfe, the co-inventor of Ethernet. And for the next two decades, they were the go-to brand for network cards. While the 3C501 network card is notoriously bad, their other cards from the 1980s and 1990s are very well regarded, and always were.
They cost more, but had good driver support, the hardware worked reliably, and they were a known quantity. And 3Com made them in the United States, which was becoming unusual even in the 1990s.
I once worked in a shop that didn’t allow network equipment unless it was one of three brands. 3Com was one of them, along with Cisco and Intel. If you plugged a no name NE2000 clone card into the network and had problems, the IT department would replace the card and bill your department. There was a saying in the 1980s that no one got fired for buying IBM. When it came to networking, one could say the same for 3Com, during their prime.
And they were huge. In 1995, 3Com bought the naming rights to Candlestick Park, the legendary baseball/football stadium in San Francisco. It held those rights until 2002.
How 3Com’s long decline happened
But after the dotcom bust, 3Com faded away. What happened to 3Com, ultimately, was a series of bad decisions that led to a loss of profitability. Some of them may have made sense at the time. But not all of them.
The first thing that happened was 3Com walked away from its enterprise networking equipment business. Admittedly Cisco was hard to compete with. But 3Com was the only brand succeeding at doing it. The logic behind the move always puzzled analysts, because it didn’t save any money. Revenue dropped, and 3Com angered its customers and dealers. It’s likely many started buying Intel cards afterward.
Many corporations like to buy multiple things from a single supplier. This was one reason IBM stayed in the PC business for several years even though they were losing money on it. The computers weren’t profitable, but it made it easier to sell services, and the services were profitable. There is always the fear that two vendors will point fingers at each other, saying the network problem was caused by the hub and the network card being mismatched brands. Interoperability was the whole idea behind Ethernet, but not everyone understood that. One could commend 3Com for not wanting to exploit that misunderstanding. But it came at a financial cost.
It also didn’t help that the company branched into other businesses outside its core. The purchase of US Robotics was a key example. US Robotics made modems and Palm Pilots. At the time, that was big business. But 3Com paid a premium price for products with a limited shelf life left in them.
Nothing good happened to 3Com after that. It took 3 years for Cisco to go from being 90% the size of 3Com to double the size of 3Com. The profitability of selling network adapters dropped. And by the time 3Com realized they needed to get back into network equipment, it was an expensive proposition. From 2000 to 2005, the company didn’t turn a profit.
Why 3Com failed
Ultimately, 3Com failed because it was too dependent on network cards. Network cards accounted for about half their revenue during their bust years. And their cards were expensive. Intel undercut their price to try to gain market share, and they succeeded.
And then they did exactly the wrong thing. It was the late ’90s, and they bought their way into the modem business. The dial-up modem business. At that point in time, everyone knew broadband was the future, they just didn’t know exactly when it was going to arrive. Yet 3Com had no plan to get into that business. It was like betting on AOL even though they were in the position to know the future belonged to AT&T.
Making matters worse, minimum viable product-style modems undercut US Robotics modems, just like the network card business. Not that Intel cards were a minimum viable product. But pretty much any ’90s network card not made by Intel or 3Com was.
3Com sought partnerships, most notably a partnership with Chinese networking giant Huawei. There were even rumors of Huawei buying 3com, though that deal never went through. And 3Com started making product overseas as a cost reducing measure. But ultimately, nothing could make up for all that lost time and all those bad decisions.
Who bought 3com
So what happened to 3Com? 3Com didn’t go out of business and they didn’t sell out to Huawei. They got Compaqed. Hewlett-Packard bought them out, which seems to be where every great ’90s technology company went to die. Hewlett-Packard paid 2.7 billion dollars in late 2009 for 3Com’s assets.
HP still sells network equipment like switches, routers, and access points. But they’re not as big of a name in networking as 3Com had been. And the 3Com pedigree isn’t the selling point. If they put a brand on it other than HP, it’s the newer Aruba brand, not 3Com. The selling point is the price being lower than Cisco. This makes them attractive to executives or purchasing agents looking to cut costs, consolidate vendors, or both. HP will usually beat Cisco’s price. And if the IT department is under pressure to reduce the number of vendors it buys from, HP sells more things than Cisco. So HP gets in the door for one of those reasons, and when their network equipment turns out to be no worse than Cisco equipment, they stay with it. At least until Dell comes around offering discounts.
But the network cards are long gone. When you buy an HP branded network card today, guess what it is? It’s an Intel card with HP’s logo engraved on it. Unless HP is trying to control costs, in which case it just has a sticker.
3Com today
Vintage 3Com network cards are very popular with retro computing enthusiasts. It had top-tier name recognition in the 80s and 90s. All but their very earliest cards generally have good driver support, work well, and are easier to configure than most of the cheaper competitors.
My favorite retro network card is a very specific Intel card, but if you need a card for anything from the 286 to the Pentium II era, you can do far worse than a 3com 3c509 ISA card or 3c905 PCI card. It’s absolutely period correct and absolutely something a knowledgeable professional would have used at the time. I have better luck with the ISA cards under DOS than the PCI cards, so if I want to run DOS and do any networking, I use a 3c509 ISA card in the system. But for Windows 95 and newer, the 3c905 PCI card works really well and is very cheap and common.
David Farquhar is a computer security professional, entrepreneur, and author. He started his career as a part-time computer technician in 1994, worked his way up to system administrator by 1997, and has specialized in vulnerability management since 2013. He invests in real estate on the side and his hobbies include O gauge trains, baseball cards, and retro computers and video games. A University of Missouri graduate, he holds CISSP and Security+ certifications. He lives in St. Louis with his family.