News Projections show that Arm CPUs will power 40% of notebooks sold in 2029

[Admin]

A semiconductor-focused market intelligence firm has charted the rise of the Arm architecture in the notebook market up to 2030. In a new research bulletin, TechInsights predicts that 40% of notebooks sold by the end of the decade will be based on SoCs using the Arm architecture

Projections show that Arm CPUs will power 40% of notebooks sold in 2029 : Read more

 

[redgarl]

No way in hell. ARM will still be marginal on Linux or Windows. ARM CPUs are buggy as hell outside of MacOS.

 

[EzzyB]

I think back to two things.

One, we've already done this. IBM bet huge on RISC in the 1990's and lost miserably. RISC was the second coming of computing, the messiah, it was going to save us all!

Two, for quite some time the top supercomputer in the world was made by Fujitsu using ARM (RISC) chips. It was actually less efficient, by a small amount, than it's x86 competitors when averaged for compute.

The success of RISC and therefore ARM, in the last couple of decades has been solely based on "good enough and uses less power." The M1 and it's successors were possible only because we then had enough overhead in processing power to make emulation at least less painful. It's "good enough and uses less power." (And even better different, making benchmarks between the two murky at best.)

That said, the power optimization of x86 is an entirely new thing. It seems you can build a 20 hour laptop with an x86 chip if you put your mind to it.

So, I'm not really ready to believe this just yet.

 

[Notton]

40% is a bold statement when the only Windows-on-ARM chip available right now is the overpriced and under-performing Snapdragon X Elite/Plus that aren't selling well.
X Elite2 is going to need something like 30% uplift in CPU and 200% GPU performance/watt if it wants to keep up with Lunar Lake and Strix Point.
If Mediatek enters the market with something that strikes a good balance between price/performance/efficiency, then sure, maybe.
Also, it requires Microsoft not constantly bungling Windows-on-ARM, as evident with x86-64 Win11 24H2.

Personally, I just want to see AMD use memory-on-package for some sweet power efficiency and memory bandwidth gains that Lunar Lake has seen.

 

[thestryker]

For this to come to fruition Microsoft would have to go all in on Windows on Arm and there would need to be more than just Qualcomm and Apple on the SoC side. There's the oft rumored MediaTek laptop/desktop chip potentially on the horizon, but I'm unaware of anything else.

 

[Kondamin]

Let’s see sub 600 usd windows laptops powered by arm before making statements like that

 

[jlake3]

I hate to claim that as an armchair expert I know more than a major market research firm... but I really feel like this is just fanciful speculation at this stage? (Or taking whatever sales projections ARM Holdings gives at face value?)

That 18% current market share for ARM appears to include Apple and Chromebooks, and most of it would seem to be Apple. If Apple's prices and compatibility keeps them mostly constant, either Windows-on-ARM or ARM versions of ChromeOS have to absolutely explode.

Right now Windows-on-ARM barely has a foothold, and a few months out from the much-hyped Snapdragon X Elite launch it seems to have already been matched by Intel on battery life and NPU and beaten on graphics while also not having compatibility concerns. More affordable X chips seem like they might not be out until after sub-$1k laptops with the new Intel chips. Mediatek's premium laptop chip is still a rumor. Nvidia getting into the laptop CPU market feels like part rumor and part wishful thinking (personally I think they seem more interested in datacenter and embedded right now, and their controlling nature not likely to mesh well with OEMs).

Over on ChromeOS.... Anecdotally, there seems to be less ARM-powered Chromebooks for sale than there used to be. The listings look to be dominated by Celerons now. My understanding is that there's less architecture lock-in on ChromeOS than Windows, so that could easily change... but less ARM SKUs does not seem like progress.

Based on who's in the laptop market, what their prices are, what their hardware and software looks like, the fact that the laptop market doesn't turn on a dime... I fully expect ARM will have more of the laptop market by 2030, but I don't see more-than-doubling unless something significant changes.

 

[dCasualGamer]

40% is a bold statement when the only Windows-on-ARM chip available right now is the overpriced and under-performing Snapdragon X Elite/Plus that aren't selling well.

They need to say things or come up with "studies" like this to build a momentum. They have to convince everyone that there is a future or else --> will developers write native codes for ARM if there is only <20% market share ? Will regular PC buyers buy an ARM machine if they do not know if the ecosystem will be there in the near future. How many systems will OEMs (Dell, HP, ASUS, Lenovo, etc) forecast to build..

So all of these are just marketing propaganda to create an illusion of inevitability. Reality is going to be difficult to predict. Five years ago, who would have guessed NVDA and INTC market cap today..

 

[Kamen Rider Blade]

Given how the SnapDragon X-Elite sales aren't doing well, I'm willing to bet on the opposite.

Eventually MS will be tired of supporting ARM, and give up again.

Then we can finally convince Qualcomm to join Team x86 =D

 

[usertests]

Yeah, Snapdragon X is a complete flop, Apple will hold steady where it is. For this "insight" to become true, the rumored MediaTek/Nvidia SoC has to land with a splash and avoid Qualcomm's mistakes.

 

[SunMaster]

I wonder who paid for this "research". It's more like paid marketing trying to gain some momentum.

 

[dalek1234]

My Projection is that whoever made the la-la-land projection mentioned in this article, will soon be institutionalized and heavily medicated (it's for his own good). Oh, and that chart talks about "shipments", not "sales". Shipments mean nothing if the laptops are just sitting on retailers' shelves. ARM laptops are good for browsing the internet and taking notes in class. Everything else doesn't fully work. Why would anybody pay so much money for something that can only do two basic tasks?

 

[dCasualGamer]

Given how the SnapDragon X-Elite sales aren't doing well,

Where can we find the sales data ?