| F E A T U R E|
The 10 Most Important Products of the Decade
Number 5: Kalpana EtherSwitch
October 2, 2000
By Robert J. Kohlhepp
Before then, bridging--a store-and-forward technology--had been dying at the hands of routing. Why EtherSwitch and not EtherBridge? Two reasons: First, bridging was a bad word in the industry, one to be avoided by any good marketer. Second, Kalpana's product didn't meet IEEE specifications for a bridge, so rather than fight, the company switched.
In the late 1980s and early 1990s, campuswide Ethernet backbones caused many enterprises to be strapped for bandwidth. As taps off the backbone were bridged into various buildings and departments, networked applications started exploding, and bandwidth was constrained.
Much to Cisco Systems' delight, most networks converted to a routed model and a faster backbone--either a collapsed backbone (using the router's backplane as the backbone of the network) or FDDI.
Enter Kalpana. After being rebuffed by various members of networking's elite, Kalpana introduced a basic seven-port Ethernet switch. The device offered very low latency, full-duplex operation, lower prices per port and, best of all, plug-and-play operation. The promise of bandwidth without the complexity of routing was Kalpana's trump card.
Kalpana's first EtherSwitch debuted for $10,500, or $1,500 per switched port. While outrageously expensive by today's standards, that price was about one-third the cost of routed Ethernet ports of the day, and management was infinitely easier. The switch didn't perform at wire speed, it didn't have high-speed uplinks, and its switching matrix wasn't particularly fast, but its backbone offered an aggregate throughput typically superior to that of Cisco's flagship router, and at a fraction of the cost. The EtherSwitch was great technology at the time and hit the market just when it was needed most.
But Kalpana stayed too long with its pure cut-through technology (transmitting the packet on the outgoing port before the entire packet had reached the incoming port), which prevented the incorporation of higher-speed uplinks.
The company was also late in addressing broadcast issues, which plagued large, complex networks and were exacerbated by the new switching technology. On the bright side, Kalpana pioneered EtherChannel, which used multiple ports on one switch to connect to multiple ports on another switch or server, thereby aggregating the bandwidth.
By 1994, Kalpana was in a dogfight with 3Com and other switching vendors, which had embraced higher-speed uplinks and had deeper pockets for development than Kalpana had.
Cisco, meanwhile, for whom routing was religion, had been growing at an incredible rate, but it had missed the switching boat, initially ignoring the "toy" devices. Switching was just a fancy name for bridging, and that was a 1980s technology--or so the thinking went. However, as the decade progressed, it became clear that if Cisco were truly going to own enterprise networking, the company would have to embrace switching--and quickly. A few jots in its legendary checkbook brought Cisco from switching sideliner to switching headliner, and one of the company's key acquisitions was Kalpana.
Kalpana's products thus lived on as the Cisco EtherSwitch line. Great innovations, such as cut-through forwarding, full-duplex Ethernet and EtherChannel, are still core technologies in Cisco's--and, for that matter, nearly all--Ethernet switching products.
At the time of the EtherSwitch's debut, Kalpana's founders couldn't have imagined the effects switching would have on all aspects of networked computing. But Kalpana's EtherSwitch, a little product backed by great vision--and great marketing--proved to be the first in a raft of product introductions from numerous companies that brought much needed bandwidth to a hungry market.