Intel should dump x86 and make ARM chips, says executive
Summary: As the PC industry stagnates, is embracing ARM the way for Intel to gain traction in the post-PC world?
There's little doubt that the PC industry is in the waning phase of its existence, and that all the players are scrabbling for new markets, but this latest suggestion from an ARM executive could be a little too far-fetched.
According to SiliconBeat's Troy Wolverton, ARM's lead mobile strategist James Bruce thinks that chip giant Intel should cast off its long relationship with the x86 architecture and instead embrace ARM's architecture. His reasoning is that Intel is struggling to gain traction in the mobile market -- Bruce claims that some estimates put Intel's share of the market at only 0.2 percent -- and that embracing ARM is a fast-track to grabbing more influence over the mobile space.
While it might be easy to dismiss these statements are bluster on the part of ARM, there is a certain logic to the idea. After all, Intel has made great investment in the mobile space, and that doesn't seem to be paying off. Also, as we move from the era of the PC into a post-PC world, Intel runs the risk of eventually sidelining itself into irrelevance. Intel continues to be heavily dependent on the PC, while ARM has squirrelled its way into every aspect of modern consumer electronics, from TVs to refrigerators.
Another threat facing Intel isn't content to stick to mobile devices, with the company putting its name to silicon that's finding its way into servers and systems designed to handle processor-intensive task.
While ARM seems well equipped to encroach on Intel's market, Intel seems ill-prepared to counter.
Is Intel likely to abandon x86 any time soon, and shift allegiances to ARM? That seems highly unlikely at this time.
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
LOL
+1
Just one quick reminder...
Intel does not push forward their architecture, but rather their finished chips. On their point of view x86 is proprietary IP and ony AMD survived the chip wars making clones, under an IP agreement on the 90's. Cyrix, TI, Mitsubishi and several other jumped wagon on the late 80's, unable to compete.
On the other hand, ARM has always pushed its architecture(s) cross-licencing it to hundreds of players. One of them used to be part of Intel, the so-called StrongARM. It came to them after the adquisition of Digital's Alpha technology division. It was used on the Apple Newton and (I think) on several Palms. Today that team is part of Marvell.
It's rather funny but the once thought ended chip wars, reemerged on mobile. DEC Alpha, MIPS, IBM's PowerPC, Sun's SPARC and HP's Precision Architecture fought hard against x86 and finally caved in. HP and Intel created Itanium as a truce. It also fell due to lack of software.
Tiny ARM retreated to the embeded camp, just like MIPS and PowerPC. Just like the prototipical iceberg, ARM grew below the radar and (re)surfaced [pun not intended] when it was mature enough.
Intel has never been a big volume player in microprocessors.
They could keep x86 on the high end and use ARM for the low end.
Multi Core
The same could be done with a 4-core Intel: replace one [or two] with ARM cores, like the cortex-A15 series.
That would roughly equate to an Intel i3, combined with an Nvidia Tegra 4.
Re: Who says all cores need be the same?
Ridiculous statement
dump is a little strong
that said though, intel would be stupid not to get into ARM with the rise of mobile. they may prove me wrong yet, but so far I'm not convinced they will ever catch ARM on power efficiency. I understand they don't want to be beholden to ARM with the requirement to license from them, but intel is digging their own hole in the mobile arena. out of all the phones at CES only 1 is intel, and will likely only see limited markets. there were no intel android tablets out of all the tablets at CES. it's clear they've failed for mobile for now, and it's only an iffy prospect that they can establish themselves there.
within the next few years MIPS is going to bring some competitive chips for high end devices too, and they have a good chance to be even more efficient than ARM is. whether that means intel should leapfrog ARM and partner with MIPS, I don't know. but it's clear x86 isn't working in mobile.
Run a server with ARM chips or what ?
Apart from that powerful servers will always run with high performance CPUs.
I'm desperately hoping that Intel will fail to get a foot into the mobile area. Intel has enough market dominance already.
Re: Run a server with ARM chips or what ?
You know what the biggest cost in running a data centre, server farm or supercomputer is these days?
It's the electricity bill.
Sounds like the ARM people are running scared that Intel might have success
To the ARM people, it's a desperate attempt to try to stop Intel before their processors make any kind of headway into the mobile markets, which would hurt the ARM providers, and might even, eventually, turn the ARMs into orphans, where people would be opting to go with the more powerful processors and the ARMs would be left abandoned.
Arm
Intel should dump x86 and make ARM chips, says executive
Yet, the "older" technology of the x86 platform, is still superior to that
In every other area, Intel's chips are far superior to any ARM chip.
So, your analogy is very flawed, and you need a new one.
Re: except in the area of battery charge consumption.
And people who run data centres, server farms and supercomputers are discovering that their biggest cost is the electricity bill. So even here there is scope for low-power technologies like ARM and MIPS.
But, you're conveniently failing to mention that, Intel processors,
Re: ARM super computer
Oh wait... how many are the supercomputer that do run Windows? :)
It seems people often forget what the "super" part means.
Changing the subject, does not invalidate what I said,
Linux is free, and thus, it's the natural selection to use when a super computer needs to have an OS installed into mullions of processors. Windows would end up being prohibitively expensive for any super-computer. But, Windows could run super-computers, but not cost-effectively.
Understand the difference, and why?
What about us dedicated decktop users?
Regards