Pioneers, arrows and the dangers of misjudging the market

But someone has to blaze a trail and, in the case of the server market, that pioneer was Calxeda. Was, because the company has closed due, apparently, to an inability to get any further funding. Investors, including ARM, will be taking a big hit; Calxeda has already 'burned' around $100million of VC cash. Calxeda – which was previously known as Smooth Stone – was pioneering the idea of using ARM based chips in servers in an attempt to bring a non Intel presence into a large and growing market. The company's chips – apparently designed into a forthcoming HP server product, but then bumped in favour of Intel – looked to bring more power efficient processing to a sector where power consumption wasn't always top of the list. Calxeda was pursuing the right market, but with the wrong technology. It was developing chips based on ARM's 32bit v7 architecture, while the market wanted 64bit processors. While ARM responded to market demand last year with the launch of its 64bit v8 architecture, Calxeda continued to develop 32bit devices. It seems business dwindled as potential customers waited for a 64bit offering. That 64bit chip wasn't likely to appear until the end of 2014 at the earliest. By then, similar devices will be on the market from the big server market players, including AMD. Potential investors will have balanced the amount of cash required to get Calxeda through the next year against the potential revenues from its 64bit devices and decided it wasn't a good bet. We talk regularly about the dangers of being late to market with products and the effect on revenues. It seems that it can be just as dangerous to be early to market.