In 2011, Microsoft (NASDAQ:MSFT) announced that Windows 8, due for release in late 2012, would be compatible with ARM (NASDAQ:ARMH) based processors, in addition to the standard x86 chips produced by Intel (NASDAQ:INTC), AMD (NYSE:AMD), and to a much lesser extent VIA Technologies. This announcement sparked a great deal of excitement among the tech community at the prospect of greatly increased processor competition, and speculation over possible outcomes. Current opinion seems to be split among 3 different scenarios, with different consequences for investors.
Outcome One: Market Segregation
In this scenario, ARM-based products and x86 based products occupy distinct market segments and have relatively little competition amongst each other. Under such a scenario, Microsoft would benefit from exposure to a larger market that is currently the domain of Apple (NASDAQ:AAPL) and Google (NASDAQ:GOOG). Hardware manufacturers should see little impact - they already serve these markets, just under a different software platform.
The key argument in favor of this outcome is that Microsoft changing the software platform does not impact the comparative advantages of the current platforms. ARM chips are generally more power efficient and lower performing than x86 chips - even the highest end ARM-based chips, while perhaps suitable for the typical user, don't compete with the performance of a modern desktop processor. The argument is that ARM is good at power efficiency, x86 is good at raw performance, and that dichotomy is unlikely to change due to Windows 8.
As an example of this dichotomy, benchmarks show Intel's single core Medfield reference platform beating the highest end ARM smartphones. Desktop x86 processors have 2-8 cores, more than twice the clock speed, a larger cache, and greater memory bandwidth. In terms of power consumption, many ARM chips are rated for well under a watt at full load, while a desktop chips are rated anywhere from 25-130w.
Outcome Two: Battle for the Bottom
In this scenario ARM based chips compete vigorously with x86 based chips at the low margin but high volume segment of the market. In addition to tablets, we could see ARM based processors moving up into thin and light laptops and budget desktop machines. For users who don't need the performance of a full on desktop processor and who care primarily about web browsing and office work ARM based machines offer a competitive alternative.
In this scenario winners would include the major manufacturers of high performance ARM-based Systems on a Chips (SoCs), including Qualcomm (NASDAQ:QCOM), Nvidia (NASDAQ:NVDA), Samsung, and Texas Instruments (NASDAQ:TXN), due to the increase in market size.
Perhaps more interestingly would be the potential harm to AMD and Intel. If you take x86 processors as a distinct market and then calculate the Herfindahl Index you could reasonably conclude that Intel has market power (in the microeconomic sense) significantly in excess of what would be expected in a perfectly competitive market. The increase in competition for Intel's high volume segment could be a substantial problem for the company.
Outcome Three: Netbooks Take Two
When Intel released the Atom processor in 2008, it sparked interest in netbooks - low performance, low cost machines. Initially users were excited about the option of a cheap home user device, and netbooks quickly became a staple item of manufacturers. Over the next 3 years, interest in netbooks began to wane as users realized the compromises inherent in meeting the price point . While the cost was appealing, users wanted greater performance than could be provided by a netbook.
There is some chance that the introduction of ARM processors could have the same effect. We may see an initial surge in interest for low costs devices, followed by a decline as consumers recognize that the low price point comes with substantial tradeoffs. In this scenario, we would see initial wins for ARM manufacturers and initial losses for Intel and AMD due to enthusiasm for the new products. This situation would latter reverse itself as the market returns to its previous equilibrium.
At this stage I don't think it is possible to make a definitive prediction about the outcome that is most likely. There are a few conclusions that can be drawn however. Intel and AMD are unlikely to come out ahead- their best case scenario is little chance in revenue, and their worst case is substantial competitive pressure on their highest volume (although not highest margin) market segment. ARM manufacturers will probably do well in the short term, but should be monitored carefully for a reversal in consumer interest. Regardless of the outcome, Windows 8 should prove a good tool for Microsoft to gain a presence in a new segment of the market.