Discussing chipmakers other than Intel and AMD in the consumer space still seems a little weird to me, having grown up largely amidst a duopoly, followed by a near complete monopoly in the consumer CPU space (less a few oddball exceptions). The consolidation we saw in the CPU space in the 2000s seemed almost untouchable for a second there, but over the last decade the landscape has been shifting beneath our feet, and fresh competition is arriving on the scene. Everyone from Apple and Google to Qualcomm are now putting pressure on AMD's and Intel's market share and profits, impacting their ability to conduct future research and development, leaving both companies scrambling to respond.
The crux of this has been ARM, which has gone from a quirky Cambridge-based designer of RISC architecture to the now dominant and defining force in the CPU space. ARM-based designs have rapidly gone from oddball competitors with poor software support, primarily designed for low power or integrated systems, to being the dominant force in mobile computing. They are now the agenda-setting alternative in cloud computing, even starting to encroach on the desktop market. If it's not obvious yet, Intel and AMD should be worried about ARM, and here are some of the reasons we think so.
Windows on Arm is more exciting than Windows itself
Windows on Arm has had an amazing 2024, which is more than I can say for Windows.
4 Momentum in the cloud
ARM has been dominating cloud computing for a while now
If you're involved with cloud computing, you'll know how dominant ARM has been in the cloud for a while now. Google Cloud Platform (GCP), Amazon Web Services (AWS), and Microsoft's Azure offer their own custom ARM-based silicon, usually at significantly reduced rates with Axiom, Graviton, and Microsoft's Cobalt 100 processors respectively. These options are often far more price competitive than their alternatives from AMD and Intel, and while Intel and ARM-based options can offer some benefits or additional features, money often talks when it comes to the cloud. Aggressive pricing has been driving companies to migrate their cloud computing needs to ARM, especially in a world where many companies are dealing with escalating cloud costs and are even considering turning back to on-premise installations for established workloads.
ARM has long been pushing its expansion into the cloud, and has enabled large cloud providers to cut their costs and improve performance using their own custom silicon based on ARM. This naturally benefits cloud providers, as these chips can be extremely efficient and there's no markup to Intel or AMD once the chip is designed. These reduced operational costs are then passed onto businesses and consumers.
We tested it: Here's how Apple's M4 compares to Intel Lunar Lake and Snapdragon X Elite
Apple's latest processor is no joke
3 Mobile dominance
Intel and AMD missed the boat for mobile a long time ago
A long-standing but often overlooked element of ARMs position is its dominance in the mobile market. The vast majority of phones and tablets are ARM powered, and the admittedly lagging but still relevant market of smart and 'Internet of Things' devices are all powered by ARM. This has long been ARMs bread and butter, making use of the low-power, high-efficiency designs that ARM chips have become known for. This dominance is creeping forward, with Windows on ARM and Apple's M-series chips edging closer to the boundary of where Intel and AMD have any relevance at all. The laptop market is poised to be revolutionized by ARM-based chips, and if the likes of Intel and AMD aren't careful, the desktop could be next - especially in markets where power efficiency on the desktop is a concern.
Intel and ARM are clearly aware of this, and have formed the x86 advisory group in October 2024 to help provide more coherent direction for the x86 as a platform, particularly to enhance the platform's competitiveness against low-power ARM chips, challenge the rise of RISC-V and react to the almost insatiable hype of AI.
The dominance of ARM on mobile and in the server space means that ARM-based designs already have a great springboard to continue to close the net on AMD and Intel, with increasing experience both in high-efficiency laptop chips and high-end server chips.
2 A culture of innovation
x86 has been scrambling
One of the long-touted benefits of ARMs flexible licensing model is the culture of innovation that it's able to breed. In recent years, we've seen this truly come into its own, as everyone from Apple to Qualcomm has been able to make use of the ARM platform to rapidly innovate, quickly break into new markets (like the laptop space), and pose a serious threat to Intel and AMD once Windows for ARM is more widely supported. The same applies to the Snapdragon X Elite, which has rapidly closed any market advantage which might have seemed insurmountable a few years ago.
While they're only sort of competing directly, Apple has been able to do the same with the M-series chips. Though the M1 seems like a long time ago now, its power efficiency and the extended battery life of later iterations have revolutionized what had really been a stale laptop market since the introduction of Intel's introduction of the Ultrabook design class in 2011.
The restrictive licensing on x86 chips makes this kind of scrappy startup innovation difficult and expensive. The consequence is that x86 development can't benefit from the same kind of market for licensable designs which RISC-V and ARM chips have benefited from.
While Intel stumbles, they've been silently making a killer integrated GPU
Integrated GPUs aren't all that great, but Intel has managed to pull off a sleeper GPU in its Lunar Lake CPUs.
1 Intel is in serious trouble
From untouchable to teetering in the space of a decade
We'd be remiss to ignore the fact that Intel seems to be in serious trouble. It's been a bad year for the chip maker, with a stock price that is tumbling, struggles to remain competitive on almost every front, and mixed reports of wider concern for the long-term health of the company. The company has been trying to develop its foundry business to transform into a US-based competitor to the likes of TSMC, but this isn't simply a top level management pivot that can be implemented in weeks or months - developing the skills and facilities necessary on U.S. soil will take years. The company has laid out its road map for 2025, but how successful it will be remains to be seen.
This culture puts Intel in a sticky spot, and may leave the company with difficult decisions about which fronts to fight on and how to prioritize. It could potentially reduce the effort the company is able to put into fresh research and development. We saw this kind of flat spin with AMD in the past, where poor decisions compound themselves, making it harder and harder to catch up with less resources. There was more to AMD's 2010 slowdown than this, of course, but the lessons to be learned are still valid.
AMD has been healthier this year, and largely isn't affected by the same problems Intel is, but we should still be looking at Intel's fall from grace as a cautionary tale.
The CPU landscape can change quickly
The CPU market is a funny thing. It can feel incredibly stagnant, almost insurmountably so. Things move slowly, take time to develop, and never happen overnight. But despite this, change seems right around the corner, with complacency, lack of competition, and an unwillingness to innovate killing even the biggest beasts. There are technical reasons for this, sometimes progress is hard for legacy manufacturers, but we've seen it time and time again (PowerPC anyone?). Intel and AMD should be seriously concerned about the rise of ARM and the opportunities it opens up for competitors, especially as both California companies are already struggling to compete.