ARM vs. x86 Battle Royal: Why and How ARM Is Pulling Ahead
ARM vs. x86 Battle Royal: Why and How ARM Is Pulling Ahead
For the last decade, ARM and x86 have been at odds. ARM attempted and failed to enter the server market; Intel, which controls the majority of the x86 market, attempted and failed to enter the cellphone market. The tablet war continues, with ARM in the lead, and the notebook war continues, with Intel largely in the lead.
Until recently, Qualcomm was primarily the standard-bearer for ARM, but that will likely change if Nvidia's bid to acquire ARM is successful. At the moment, Nvidia appears to be more interested in servers than smartphones. They do, however, have their sights set on PCs, where they are likely to run into conflict with Qualcomm.
However, this is a platform war, and ARM currently has a stronger developer community than x86. This is primarily a battle of business models, with ARM focusing heavily on licensing and x86 on sales.
Let's discuss how this battle is likely to unfold, as the outcome is far from certain and the victor may end up holding all of the cards. We'll then conclude with my product of the week, a new Plantronics headset that may be just the ticket for your next Zoom call.
Old School vs. New School
The market appreciates diversity of vendors, but not diversity of architectures, as this introduces selection risk.
OEMs and IT organizations collectively invest billions of dollars in technology, and the more viable types of technology available, the more likely one of these large, influential organizations will guess incorrectly.
They'd prefer two redundant vendors who are plug-compatible, which would allow them to switch vendors later in the manufacturing process and mitigate supply risk.
The x86 had a significant advantage at one point because when IBM licensed their x86 technology from Intel, they demanded that Intel license it to another company, which turned out to be AMD, to ensure competitive pricing and backup in the event of a chip shortage.
Intel, on the other hand, negotiated socket compatibility away in order to lock down the platforms they built, unaware that this would strategically weaken x86.
Thus, Intel deviated from market expectations in order to prevent AMD (which had been growing) from gaining market share. However, they weakened x86 strategically in the process. Or, to put it another way, at a time when the computing market was shifting toward a licensing and compatibility model, Intel took the opposite path, putting both its technology and the company at risk.
In some ways, just as Microsoft's errors in the 1990s largely drove Linux, Intel's errors in the same decade increasingly drove ARM. Microsoft responded to their threat by increasing its openness and closing the gap on Linux, and even by adopting the competing platform. Redmond's defense evolved from one of strength and force to one of cooperation and advancement, and Linux is now both viable and a minor threat to Microsoft.
The market appreciates the open-source collaborative model that surrounds ARM, as well as the broader selection of ARM vendors, despite Qualcomm's arguably dominant position in ARM, similar to Intel's in x86.
This structure enables the ARM ecosystem to function more cohesively in comparison to x86. Intel, AMD, and VIA rarely act cooperatively and are more likely to ignore or position themselves against one another than they are to act in concert against the broader threat. Often, it appears as though the x86 factions' infighting diverts their attention away from the more serious threat posed by ARM.
This new competition is less about technologies and more about business practices, with ARM meeting the stated needs of OEMs and cloud providers who value both choice and a deeper relationship with the companies that build the technologies they resell and use.
Power Balance
The two sides are relatively evenly weighted in terms of power, with x86 dominating in PCs and servers, which typically have relatively high margins; and ARM dominating in smartphones, tablets, and, more likely, embedded in appliances and IoT devices.
This competitive dynamic enables ARM to maintain a commanding lead in higher-volume categories and to benefit from economies of scale. However, x86's stranglehold on PCs and servers is much more difficult to break, which somewhat offsets the potential cost advantage that would ensure ARM's eventual victory.
Additionally, x86 has a greater performance headroom than ARM, which mitigates the use of virtual machines as a leveling agent, which would otherwise provide a more direct path to competitive displacement. However, it is far easier to increase performance while reducing energy consumption than it is to maintain performance while reducing energy consumption.
Likewise, Intel ceased operations of its Developer Forum. In comparison, ARM developer events continued to shift developer support between the two platforms, significantly weakening x86 (they are correcting this, but many of the developers have already switched).
Also, with Apple's shift to ARM and Nvidia's acquisition of ARM, two of the most powerful companies in technology have effectively switched sides. Nvidia also has a sizable server beachhead, with its targeted GPU solutions for AI and analytics workloads serving as an onramp to its upcoming blended ARM server solutions.
Fortunately for Intel, Apple does not collaborate well with anyone, limiting the company's contribution to the ARM ecosystem to optics – though optics can be extremely powerful in this case.
In general, the long-term advantage has shifted to ARM. If the ARM providers cooperate, provide a seamless way for their processors to communicate with one another (the equivalent of socket compatibility), and mitigate attacks against one another, this could be their game to lose. They have not yet arrived.
If the x86 providers can band together to defend their shared platform and transition to a more open source and licensing model, as the market appears to prefer, they will consolidate their hold on their segments and be in a much better position to defend their positions. That, too, is not occurring.
Thus, this is still a game that anyone can win or lose. However, with Apple and Nvidia's moves, the momentum is clearly on ARM's side at the moment.
Wrapping Up
We are witnessing a showdown between two hardware architectures: x86 and ARM.
Prior to Apple and Nvidia effectively switching from x86 to ARM, resources and capabilities were comparable. As a result, ARM has the advantage; however, a lack of cooperation and coordination harms both sides in this conflict, ensuring that neither can fight at full strength.
Intel's efforts to rebuild its relationship with developers will take time, particularly if those developers continue to view x86 as a declining platform. Given Apple's closed nature, it may be more of a liability than an asset for ARM.
The outcome of this battle, at least for the next decade, is likely to hinge more on Nvidia's execution than on any other player, because Intel is constrained by past mistakes; and it is Nvidia's acquisition of ARM that could have the most significant impact on this battle in the near term.
However, the market desires a single technology, and ARM's open nature, which aligns with the open initiatives of software and cloud providers, provides it with an advantage that Intel and x86 have yet to fully address.
IBM and Microsoft have demonstrated Intel's path to victory, or at the very least a stalemate. However, shifting from Intel's current hardware model to one that more closely resembles Qualcomm's hardware model may be beyond Intel's willingness to change.
As is the case with most battles, whoever wins this one will likely be the one that best assesses the battlefield and can position its resources most effectively against their competitive threat, including partners and licensees.
Courses and Certification
Basic Electronics Course and Certificate
Artificial Intelligence Course and Certificate
Computer Programming Course and Certificate
Computer Engineering Course and Certificate
