ARM Holdings made big waves a few years ago when it announced it would develop an enterprise-class server version of the architecture that has taken firm hold of the mobile and smartphone markets.
But while the company has released architectures that support both 32-bit and 64-bit server operating systems, and has even seen a number of high-profile deployments like PayPal, it is fair to say that few enterprises have gravitated toward ARM processing in a significant way.
That doesn’t mean ARM has given up on the enterprise, however. In fact, it’s just the opposite. The company told analysts recently that it hopes to capture 25 percent of the server market, duplicating the gains it made in both the mobile and embedded markets earlier in the decade. The challenge, as Motley Fool points out, is that neither of these markets was dominated by an entrenched player like Intel, and the company has already seen some setbacks with the losses of Calxeda, Nvidia and Samsung as server partners. On the upside, though, the company has a killer R&D program and is getting software support from HP, Red Hat, Facebook and other top players.
ARM, of course, does not make its own chips – it just creates the architecture. So it needs actual devices in the field to make any real headway, and on that score it is getting support from companies like Scaleway, which is using the quad-core Marvell ARMv7 for a line of microservers that will be primarily deployed by its parent company, France’s Online.net, for hosting support. The company also has a foot in the door at HP, which is using the ARMv8 chip for some of its Moonshot systems, like the Proliant m400.
But with all of this activity, why are we still waiting for the big ARM server deal that will prove once and for all that the technology is capable of competing with the all-mighty Xeon, asks The Platform’s Timothy Prickett Morgan. Most people, even long-term users of a particular product or service, welcome competition, if only to keep their preferred solution provider on the ball. With AMD struggling to make a significant dent in enterprise-class hardware, most observers have been waiting for ARM to fill the void for more than two years now. The problem with developing a viable ARM server isn’t with the hardware, however, but the software, and even though the 64-bit ARMv8 has been out for four years, we are still waiting for a credible implementation for enterprise-class workloads.
This is part of the risk inherent in the business model that ARM has chosen. By sticking to the design of the core architecture and leaving the rest to the industry at large, the company has given up a lot of the leverage needed to push a product from the drafting table to production environments – to get the ball over the goal line, so to speak.
As I mentioned above, ARM has the luxury of core market domination to fund the research and development needed to penetrate new fields like servers and networking, but since these bread-and-butter revenue streams are coming from largely saturated markets, it is unclear how much of an effort ARM is willing to give to the enterprise unless it starts to produce real results soon.
Profit is the ultimate motivation, of course, so as long as the enterprise market holds out hope for a substantial upswing in ARM’s fortunes, the company would be foolish to give up at this point. The only question is whether potential revenues will be enough to offset development costs, and whether a killer ARM server design can be readied before another architecture comes along to propel the enterprise even further into 21st Century infrastructure.
Arthur Cole writes about infrastructure for IT Business Edge. Cole has been covering the high-tech media and computing industries for more than 20 years, having served as editor of TV Technology, Video Technology News, Internet News and Multimedia Weekly. His contributions have appeared in Communications Today and Enterprise Networking Planet and as web content for numerous high-tech clients like TwinStrata and Carpathia. Follow Art on Twitter @acole602.