Last week, Cisco reported a major component failure with its new suicidal server blades. Evidently, the company tried a software fix first and another server flamed out after the patch, so it decided to do a recall. These issues were common in the early PC years and the related costs forced the existing vendors to rethink both their testing and supply chain (the cheapest vendor was no longer the best) and we haven't seen failures like these from them for a while as a result.
This showcases one of the problems associated with a new vendor, even an experienced enterprise vendor, entering a new segment. While the vendor can often acquire competitive technology, it is the experience of decades that can lead the company to repeat the mistakes of others and this does represent a risk you should anticipate.
Understanding the Cause
I have been called in to help understand why this happens a number of times and over the last several decades I've seen a pattern develop. At the core of the issue are a series of organizational problems that few seem to grasp at the outset of a move into a new technology area. The very first is a lack of institutional knowledge for the segment.
When a firm views another market from the outside, it tends to see that market and diffract the image through the lens of its own experiences. A server vendor would see a switch as if it were a specialized server and a switch vendor would see a server as if it were a specialized switch. However, even though both products can be rack-mounted and look really similar, at least in a glass house, they are dramatically different inside.
Both products have processors, memory and internal logic and are surrounded by software and firmware, so it is easy to fall into this trap. Folks initially thought there was a huge similarity between telephony and networking technology and computers, leading IBM to buy ROLM (later sold) and AT&T to buy NCR (later reversed) several decades ago - both represented the largest mistakes either company made in that period and AT&T contributed to that company's failure. ROLM no longer exists and, on paper, likely remains the largest financial mistake in IBM's history.