ARM Doesn't See Intel as a Competitive Threat

[kronos_cornelius]

If it is a betting game, my money is on APU ARM's. The only way Intel can compete is if they start by forfeiting the architectures that gives them competitive advantage. In short, they should develop an x86 incompatible mobile chip.

 

[alidan]

amd wasn't scared because intel was FAR under estimating what their cpu could do.

[citation][nom]James296[/nom]and this is they're first mistake, underestimating your enemy.[/citation]

they understand their enemy clearly. intel is new to the tablet and phone space, and have no easy to move code to their processor, and it will most likley cost more than the arm equivilant, and in most cases, phone wise, speed is good enough, so price and power are bigger concerns. arms chips are also improving faster than intels also, making arm a bigger threat to intel than intel to arm.

you also have to underatstand that almost everything that has a cpu in it, is using an arm design. even if the high end phone segment got taken by intel, arm has plenty to fall back on.

but intel on the other hand can barely compete with arm on power draw, and most people dont need more than a netbook power anyway, so i could see a very large and cheap arm laptop/tablet laptop segment open up.

[citation][nom]Kelvinty[/nom]Intel has "too much" resources, and is actively trying to create many more enemies (competitions), it might be wise in the long term, but it could also be very bad news if something horrible come on, e.g. being gang banged by the "alliance" of competitions or even intel's own fk up."New" / "Newer" projects, SSD (potential new controller), Ultra"fail"book, mobile chips, "somewhat" still cannot be improved graphics, new CPU architecture, and long term advance material research...[/citation]

the newest gpu demos from intel are pretty impressive, i dont know what settings, but i hear they are playing rage very well. they may have caught up with the integrated. at least till amds next gen.

[citation][nom]Rantoc[/nom]And another company that underestimates Intel at its core business - Historically resulting in that Intel will put in the overdrive and make a huge leap (Core2 thanks to Amd's underestimation for example) and now it would seem it is about to happens again (this time the leap is thanks to Arm).As omega21xx said, the next gen of phone cpu's from Intel will be way more interesting even when the current was able to beat the current competition in most benchmark without drawing loads of power. The next atom get the benefits of Intel's advanced manufacturing and likely several tweaks to improve efficiency. The biggest question fis if Intel will use the extra headroom from the shrinkage ect to boost performance or lower the already in line battery requirements even further.[/citation]

arm has a fairly steady road map planed out, where they are gaining 30-50% more cpu power a year (i read it here, you may be able to find the article) where as intel is only making about 20% cpu jumps if that.

intels first chip will suck. its adoption rate and how big of an ap segment it can get will deturman how much better its 2gen attempt will be.

[citation][nom]NuclearShadow[/nom]Oh ARM how foolish can you get? Buy stocks in Intel now because soon they will have a stranglehold on the mobile CPU market. I for one want a piece of that pie.[/citation]

their first cpu will burn, get in on the stocks about half way though its cycle and before the 2nd gen comes out.

 

[ta152h]

[citation][nom]KardisF1[/nom]AMD had been in a superior position (albeit they pulled ahead of Intel instead of starting out ahead) during the P4 days, and they lost that to Intel. ARM's architecture is more towards lower power consumption, where x86 is geared toward raw strength. I think that optimizing for power consumption will be a relatively easy task for Intel, given their R&D. Lastly, ARM doesn't manufacture their own CPU's, so they have less control of what's on the market.[/citation]


Wrong answer. x86 sucks at everything. It's a bad instruction set, that is used entirely for compatibility, not for "brute strength". It has a performance handicap, as well as a power and size handicap.

Intel should just go with a fresh instruction set and they would be better off, because compatibility isn't that important in a smart phone. On the desktop, yes, but on a phone, it's hard to understand why it's so important. The costs for it are though.

 

[Guest]

Given the track record of both camps. The ARM camp is much more confident and flexible then the Intel
camp (given Intel's poor record at mobile devices, power consumption, and their all in one fit x86).

 

[blazorthon]

[citation][nom]ta152h[/nom]Wrong answer. x86 sucks at everything. It's a bad instruction set, that is used entirely for compatibility, not for "brute strength". It has a performance handicap, as well as a power and size handicap. Intel should just go with a fresh instruction set and they would be better off, because compatibility isn't that important in a smart phone. On the desktop, yes, but on a phone, it's hard to understand why it's so important. The costs for it are though.[/citation]

I have to disagree. x86 can be just as good as ARM in the mobile market if utilized properly and it really isn't a bad architecture. The problem with it as I see it is that we have plateaued in per thread parallelism and multi-threaded code is difficult to write, if not impossible for some tasks. x86 can't solve this the way it is so yes Intel should use a different arch or make some serious changes to x86 but as of right now x86 still has more performance head room, especially with Intel's current Medfield smartphone Atom chip. Further die shrinks can be done with x86 and it still has a ways to go before it becomes unfeasible to continue. x86 will have much trouble getting more parallel and each nm node will provide greater diminishing returns but we might not have any serious problems until well after we go below 10nm anyway so I don't think Intel has a need for better architectures. When that time comes I don't think ARM will overtake x86 but very different ISAs like EDGE and NISC will reign supreme.

Watch the TRIPS, WaveScalar, and other so-called successors to VLIW (many ISAs are considered this. NISC, EDGE, and EPIC are a few. EPIC has thus far shown itself to be weak but the other two have not) to see what happens once x86 truly runs out of steam.

To be honest if I was in control of Intel I would like to have a better architecture being worked on right now so when the time comes for it it will be ready as soon as it becomes necessary.

 

[danwat1234]

[citation][nom]Omega21XX[/nom]By the time the intel phones show up on the market, intel will be behind by a small margin again. Intel can compete but their first chips won't take much from ARM. I will however be interested in intel's next chips, this one not so much, app compatibility is a big problem too.[/citation]
Intel will most definitely be on the heels with lithography. 32nm this year, 22nm next year, 14 or 16nm the year after that.. Should be interesting

 

[neiroatopelcc]

[citation][nom]ta152h[/nom]Wrong answer. x86 sucks at everything. It's a bad instruction set, that is used entirely for compatibility, not for "brute strength". It has a performance handicap, as well as a power and size handicap. Intel should just go with a fresh instruction set and they would be better off, because compatibility isn't that important in a smart phone. On the desktop, yes, but on a phone, it's hard to understand why it's so important. The costs for it are though.[/citation]
While x86 compatibility mightn't be as important, I think it's still going to be the only thing Intel can sell their idea on.
Both arm and via are vastly better at performing using very low power. Intel never understood that wattage was measured in decimals for such devices, so I don't expect them to be very battery friendly. They'll instead have to gain marked share by providing an easier time for developers porting their x86 software over to a tiny platform.
In any event we'll have to see how it fares.

 

[Guest]

Up to now Intel was not able to feel the big deals such as mobiles development as Apple was, always somebody get the way and Intel enter afterward in with all effective... lack of mobility and dynamism..

 

[rantoc]

[citation][nom]king_maliken[/nom]Well I hope this is just a face for the public, because overconfidence is never a good thing. Well here's to hopping for a solid competition between the two companies for years to come.[/citation]

I would believe its more to keep everyone calm, because when the investors/shareholders see the competition gaining they get cold feet and that could be the end of the company (and the ceo salery) so seeing the reason for bending the truth is easy to see

 

[dickcheney]

R.I.P. ARM

 

[coupe]

Given the assets and experience that Intel has, I would be very worried when they look anywhere near you.

 

[ojas]

[citation][nom]givemeabreak[/nom]southernshark, Medfield is just chip 1, there will be a lot more to come in 2013 when Intel updates the ATOM architecture and goes all 22nm Tri-gate on ARM. The competition is just beginning.[/citation]
In 2013 Intel may directly launch an Atom using their 14nm Haswell arch. That would be fun.

 

[coupe]

[citation][nom]Zingam[/nom]Way too many Intel indoctrinated zombies around here...[/citation]

I'm an AMD fanboy and still you can't deny the power Intel has! They are THE dominant force in microprocessors.

 

[cookoy]

And a few years into the future, the executives of ARM looking back, regretted: oops, we thought intel was only joking when they said they're going to take over our market.

 

[rohitbaran]

I say good luck ARM. You will need it.

 

[atikkur]

yeah.. it would be cool to run all my programming to x86 based phone... dream comes true.

 

[Niva]

The comments on this article crack me up. At this point Intel is not a real threat, but a potential one. If Intel manages to pull off a stand in the mobile CPU market and even dominates in the cell phone/tablet market in the next couple of years, Arm still has business elsewhere which is huge. Actually Arm is a real threat to Intel, the fact Intel is now making inroads into an industry which has been dominated by Arm is a response to this threat. Even MS is now making their next OS compatible with Arm. The world is changing rapidly, and while I have utmost confidence in Intel's ability to execute well in this market, it still has to do it before Arm responds directly.

In the meantime Arm has made much more dramatic improvements in their top CPUs than Intel has in the last 5 years. If anything Arm is on a much faster track of improving than Intel. x86 architecture has serious drawbacks, and Arm has reached that point where they are actually competitive in terms of price/power and performance. It's all very scary for Intel and AMD.

 

[fidgewinkle]

When evaluating a competitive threat, the first step is to look at what stands in the way of a new entrant into a market. Looking at the mobile space, there are at least 4 different competing software markets on three different OSs, none of which is more than a few years old or contains code that takes more than a few man-months of development. Further, most of the software runs on an interpreter and is hardware agnostic. This seems to be a wide open market with almost no barrier to entry. Almost any quality product with decent backing has a chance. The opposite would be the x86 computing market where all of the software is compiled and optimized for x86, a lot of it serves a small niche and often involves a much longer development effort. This is the very definition of a barrier to entry. Thinking that an ARM interloper is going to just step in and take significant market share is sheer lunacy. There would have to be an unbelievable performance advantage, which just doesn't exist. There will be some tinkering with the idea of ARM in place of current x86, but it is a really hard path compared to breaking into mobile.

 

[Guest]

Interesting comments but missing something fundamental. Look at how much ARM charges per core compared to INTEL. The markets ARM are in, are price sensitive and also power consumption very important. Not power.
Windows 8 comes out...it be interesting to see price differences between INTEL vs. ARM on tablets.
First real competition for INTEL.
I see INTEL losing market share to ARM.