Hacker News Reader

Comment by taylodl | Hacker News Reader

Is Intel's market essentially restricted to the server space these days? Laptops and Mobile devices are (or are transitioning to in the Windows world) ARM processors, i.e. not Intel. AI is "owned" by NVIDIA. Of course, that could change but Intel doesn't seem to be interested in the AI market. Doesn't that just leave them with servers? As Windows transitions to ARM, and Linux is already available on ARM - how long will the server business last?

Are we witnessing the end of Intel? I just don't see what future they have outside of niche markets that don't support current revenues.

What's the consensus on HN?

View on HN

The number of non-Apple desktops and laptops running ARM is likely a rounding error, and that's unlikely to change anytime soon. Intel has never had a significant presence in mobile devices in the current smartphone era, and they've survived thus far.

Intel does have a presence in servers, but AMD has been offering vastly superior products for years in that space, and consequently, AMD has significantly eaten into Intel's server market share. The hyperscalers are also investing in ARM server products, but it's unclear what effect this has on the market.

I think we're witnessing the end of Intel as we know it, but not due to ARM. Rather, the current CEO -- Pat Gelsinger -- has bet the company on Intel becoming a foundry for other chip designers. Whether that bet will succeed is up in the air.

My 10-year prediction is that Intel will survive as a foundry -- semiconductor manufacturing is a national security priority and the US government can't afford to let Intel fail -- but the chip design segments of Intel may end up getting spun off into their own company (the inverse of how AMD spun off its fabs into what is now Global Foundries).

Laptops are starting to transition to ARM processors. The process is only beginning to begin; it hasn't really gotten going yet.

I agree about mobile. I agree about servers being in danger over the long haul.

The x86 architecture has been the winner by default for decades. That is still mostly true, but it looks like it may no longer be true fairly soon.

If CPUs become commodity chips, Intel may not have the financial strength to continue to be both a chip design company and a fab company. It may split off the fab business. (This is speculation, not really a prediction.)

Intel has the Gaudi 3 to address the AI market.

https://habana.ai/products/gaudi3/

Intel still has a large part of x86 market share.

They are bleeding money because they are trying to catch up in the Fab business. If they succeed with fabs, they should be very profitable again.

At the moment even they use TSMC to produce their best chips.

What's the future of the x86 market with Microsoft developing a version of Windows to run an ARM? That's what makes me think Intel is on the verge of losing the desktop/laptop market. They may still retain the server market, but as more businesses transition to cloud-based servers, the cloud providers are going to be interested in developing their own ARM-based SoCs to better manage their acquisition costs and decrease their power costs.

If Intel is planning on using their Fabs to make non-Intel devices, then I can see lots of growth in that area. Since TSMC is fumbling with their fab in Arizona, then perhaps Intel can pick up some of that business?

Windows on ARM is over a decade old

https://en.m.wikipedia.org/wiki/Windows_RT

But it wasn't a full-fledged version of Windows though, right? Isn't what's going on now is Microsoft is developing a full-fledged version of Windows to run on ARM?

What current Intel products are fabbed at TSMC?

I'm having to buy a laptop for a family member and the software they use (DAWs) are not yet compatible with ARM. So it's AMD or Intel of which there are hundreds to choose from.

I was hoping ARM since I hear battery life seems to be better but doesn't seem like it's industry-wide at all.

Reply on news.ycombinator.com