https://www.tomshardware.com/pc-components/cpus/x86-reigns-supreme-as-snapdragon-x-elite-chips-captured-just-0-8-percent-of-the-market-with-720-000-units-sold-in-q3-2024-qualcomm-misses-out-on-rising-ai-pc-sales-with-intel-and-amd-taking-charge
10 replies (most recent on top)
@sas+1vKS6oks
You do realize that Qualcomm acquired Nuvia in 2021, which was an ex-Apple employee startup...
Even so, x86 belongs to AMD now.
Old products take longer to die than you forecast...
but a slow death can be more painful then imagined.
It appears to me that Msft has long since become architecture agnostic, and if anything is already all-in on ARM.
Makes little sense for them to be tied to proprietary architecture like x86.
They will support it for a long time to come, till the market share gets too low.
Intel and AMD can continue to build great products on x86, but the other vendors can't even use that architecture and RISC (ARM and RISC-V, and any other RISC) is easier to develop.
The industry is close to a decade past the point where the extreme complexities of x86 were needed to be competitive. That's why x86 is losing market share, and why proprietary architecture always loses out to open source over time.
Pat sees that and by breaking out Foundry he is setting the company up to grow what will be mostly ARM manufacturing. That capability then would set up the existing or a new product division to pick up ARM (again) or even move to RISC-V.
The company has in the past shown no capability to stick with low profit margin (or low volume) products, so moving to RISC-V any time soon seems unlikely.
The real question should be: Will Microsoft be on top forever? If the answer is no, start moving away from X86 ASAP.
It's not about where the market is today, but about where it is headed.
Proprietary architecture always loses out to open architecture.
It is inevitable, but of course x86 still has most of the PC and datacenter market share for now.
That makes x86 a target and 2 companies up against the rest of the industry is a losing battle.
Same way that NVDA will inevitably lose market share over time. CUDA will entrap them if they hold on to it too long, no different than how x86 has trapped Intel.
Look at all the prior attempts to develop ARM products, that all got shelved because of the x86 profit margins.
AI ARM PCs are a Microsoft funded thing.
They threw ungodly amounts of money to QCOM and OEMs.
Like most consumer products, MS has no idea what people actually want.
This shows why an engineer would want to work for Qcom over Intel. Qcom is always on their feet. They know they can't rely on mobile and are trying to expand their presence in other areas. Before it was servers. They've failed, but it creates excitement and innovation within the company.
OTOH, Apple came to Intel to ask to make the CPU for the iPhone and Intel declined because they thought it was "low volume." The rebuffed Sony in a similar manner. Now Intel is trying to play catch up in AI/GPUs.
The only viable Arm PC is the Mac … Period, they are wasting their time and money trying.
Even if they did work properly , they would have to be so cheap that there is no margin in it for the retailers or oems … literally pointless.
So I’m sure x86 will dominate the laptop space for years to come … risk for Intel , is that people continue to buy laptops period … we can see now, iPad or old PCs are capable of what most people need.
Current AI PC ads running are a total joke , and will not convince many …
AMD has better products.
Let’s see what Nvidia will do in this space.