Wall Street's PC Fixation Blinds It To Strategic Pivots By Intel, Microsoft

It's a tiresome routine. Every quarter Wall Street analysts calibrate their earnings estimates for Intel , Microsoft , Micron and other 'old tech' companies according to the latest IDC and Gartner PC sales figures. PC sales down? Intel and Microsoft are sure to take a beating. It's as if these are still monolithic businesses stuck in 1995. Yes, Intel did have to reset expectations a month before the earnings call due to soft PC sales, but the WinTel collective has moved well past its dependence on Windows licensing and laptop Pentium shipments. Similar stories play out at other firms that grew out of the PC era, where NVIDIA , with its deep learning and autonomous vehicle efforts, Micron, via enterprise SSDs and mobile memory, and, of course, Apple have pivoted to new businesses or invented entirely new product categories. Still others, like AMD have been left behind as alternative suppliers like ARM, whose revenues rose 14% in the last quarter, capitalize on generation mobile.

Indeed, one of the most interesting and significant trends in the tech industry is the adaptation of established, old tech stalwarts to the post-PC world defined by mobile, social, cloud and big data analytics. For the most successful, PC sales, or associated client-server enterprise infrastructure, are much less important to their strategies and long-term viability. For others, like Dell, EMC and HP, transformative creative destruction is a work in progress.

Intel: Inside Your Data Center

Intel's recent first-quarter earnings release clearly demonstrated that the company is much less dependent on the vagaries of the secularly declining PC market than many give it credit for. Expect similar strength from Microsoft's commercial business, particularly its cloud services, when it reports on the 23rd as it too has become much less dependent on the PC sales cycle. Unlike former WinTel competitor AMD, which issued a disastrous report on the 16th that lopped 14% off its stock price, Intel, Microsoft and other stalwarts of the PC era are successfully executing large, complex strategic transformations to a world of smartphones, connected devices and cloud services.

While still playing catch up in the mobile device market, Intel's focus is in on the  supporting cloud infrastructure and its Data Center Group was the star of the latest quarterly report. Although the Client Computing Group (PC CPUs and chipsets) still accounts for almost 60% of the total revenue, Intel's Xeon line and other infrastructure hardware actually generated more profit in Q1: 65% of total operating income. For the past 12 months, income from data center products increased 35% versus 18% at a client group helped by PC upgrades by companies replacing now-unsupported Windows XP systems. Indeed, data center revenue rose 19% last quarter with ASPs up an impressive 5%. The relative strength of Intel's infrastructure results will continue, if not accelerate, as it has a strong pipeline of Haswell (v3) server chips, Broadwell-based SoCs and new, high-capacity 3D NAND flash devices. As I wrote last year, all are aimed at winning share in all corners of the data center. With work- and face-time moving from PCs to mobile devices, and the concomitant migration of enterprise workloads to cloud services, aside from its persistent mobile struggles, Intel is well positioned.

A standard chart in most Intel data center strategy slide decks explains why: the correlation between device growth and infrastructure capacity. On average, every 400 smartphones require one server's worth of infrastructure, whether for cloud file sharing, email/messaging service, wireless network management or mobile app backends. The ratio is even lower (i.e. more favorable for higher server spending) for wearables, connected industrial equipment and digital signage. None of these client devices may actually use Intel chips, although the company is working hard to address shortcomings in its low-power embedded and mobile offerings, however given Intel's data center dominance, including a virtual monopoly in scale out cloud servers, the company still wins by selling fewer, higher margin server CPUs, chipsets, communications processors and solid state storage.

More to Microsoft than PCs

A similar transition from client to server happened at Microsoft where revenue from commercial licensing, things like Office Pro, Exchange, SharePoint and SQL Server, accounts for almost half of total revenue and over 60% of gross margin. Yet Microsoft's fastest growing business is commercial cloud services — Office 365, Azure, and Dynamics CRM Online — where the current quarter's run rate should significantly exceed the $5.5 billion reported in Q1 as businesses shift to cloud services. One indicator is an interesting fact that CEO Nadella noted in Q1: 45% of Office renewals went to the cloud. [see update below from Microsoft's April 23rd earnings release]

Microsoft's Azure has emerged as a significant competitor to AWS for cloud infrastructure and is particularly popular with legacy enterprise customers who see Azure as the logical choice for building hybrid clouds. As I pointed out in covering the Rightscale cloud usage survey, Azure IaaS use has doubled in the past year and Rightscale's data show it soon challenging OpenStack for the number two private cloud stack behind VMware. Even Windows 10 has been retooled, in Nadella's words, "for the mobile-first, cloud-first world," although as I argued last week, Microsoft could go one better by building a cloud-native thin client.

If Intel and Microsoft illustrate companies successfully executing the post-PC, mobile-cloud transition, AMD is an example of one that didn't. Although the company has made plenty of noise about low-power, hyperscale servers tuned for cloud customers, its ARM-based road map just slipped a full year and it effectively admitted defeat in the server business by announcing plans to jettison SeaMicro. Just three years after paying over $330 million for the self-described "pioneer in energy-efficient, high-bandwidth microservers," AMD took a $75 million write off of intangible assets (i.e. goodwill). Of course, since the turn of the century, AMD has an aggregate operating loss of over $3.5 billion, so 8- and 9-figure charges are nothing new. Although the company has had success in the game console market, that has barely stanched the bleeding as it lost mainstream PCs to embedded Intel graphics chips and high-end gamers to NVIDIA, which now owns 75% of the discrete GPU market.

Once a legitimate challenger in the x86 server business, AMD lost its technical edge and hyperscale servers were its one chance of riding a disruptive technology to success. However the company did little with the talent and technology acquired from SeaMicro, let others like AMCC take the lead in 64-bit ARM server processors and now faces existential threats without a clear and compelling strategy. AMDs upcoming analyst day better provide more answers than questions.

The technology companies that have survived and ultimately thrived after the combined financial/technological vortex unleashed a decade ago have been adept at creative destruction: unafraid to cannibalize existing businesses to gain early mover advantages in emerging technologies and markets. Those that hesitated, like Intel and Microsoft, trying to protect established cash cows, are still recovering ground lost to more adept rivals. Those that dithered, like AMD, BlackBerry, HP or even RadioShack, face difficult and sometimes futile fights for survival.