ARM has declared its intentions to dominate the future of computing, declaring industry giant Intel a dead duck due to a misunderstanding of what customers want.
The comments came from ARM co-founder Hermann Hauser during an interview with The Wall Street Journal, in which he declares that 'ARM is going to kill the microprocessor.'
Hauser goes on to clarify his remarks: 'the reason why ARM is going to kill the microprocessor is not because Intel will not eventually produce an Atom that might be as good as an ARM, but because Intel has the wrong business model. People in the mobile phone architecture do not buy microprocessors. So if you sell microprocessors you have the wrong model. They license them.'
That distinction between Intel, which designs and manufacturers its processors, and ARM, which designs the processors for other companies such as Texas Instruments, Marvell, and Samsung to licence, refine, and build themselves, redefines the battle, says Hauser: 'its not Intel versus ARM, it is Intel versus every single semiconductor company in the world.'
Hauser is clearly betting heavily on ARM to take over the PC industry, claiming that 'there is no case in the history of computing where a company that has dominated one wave has dominated the next wave and there is no case where a new wave did not kill the previous wave - as in obliterate them,' declaring the era of Microsoft and Intel to be drawing to a close.
It's hard to argue with ARM's corporate performance: this year the company has collected more revenue from its licensed designs than Intel has on its microprocessor sales, while still allowing its customers to make a profit of their own from the chips they manufacture. Increasing interest in the low-power chips from netbook, ultra-portable, and even server manufacturers shows that ARM's long absence from the desktop and server markets could be drawing to a close.
Intel, for its part, unsurprisingly disagrees with Hauser's comments, claiming that 'there's room for many [different architectures] to be successful,' but it's clear that the company is rattled by the potential of ARM's designs - and this is the first time in many years that Intel, the giant of the server world, has entered a battle as the underdog.
Do you agree with Hauser that the sun is setting on traditional microprocessor manufacturers and that the future belongs to ARM, or would you need to see a high-end ARM design with the same performance as one of Intel's Xeon chips before you gave up on the x86 giant? Share your thoughts over in the forums.
The software infrastructure has a long way to go on ARM before can get anywhere near to rivaling WinTel.
ChromeOS / Linux on ARM is not going to put much of a dent in WinTel in the desktop space.
Chrome OS cannot run standalone applications, unless they changed something. It boots to a web browser.
It sits on Linux. All any Linux app needs to run is a wrapper so that it can sit inside of Chrome. It will take a bit more coding to adapt, but it is going to be tinkered with anyway to optimize for ARM. And of course it will need to be prepped through some type of marketplace.
Processor intellectual property licensor ARM Holdings plc is expected to reveal its plans for processor cores that support 64-bit computing, within the next few weeks, according to an IDG news service report that cited sources close the company speaking at an ARM technology conference in Taipei.
ARM (Cambridge, England) has shown "samples" according to the report (see below).
The move is not unexpected although the detail of how a move to 64-bits would be archieved, its multicore support and when production volume chips would be available, would be of great interest.
ARM is already known to be working with a number of chip and equipment companies on applications of its cores within server applications but one thing the company has lacked is the ability to process data 64-bits at a time, which is a standard approach in the mainstream server and supercomputer markets.
A speedy move to 64-bits would show that ARM is serious about its desire to compete with Intel Corp. on the chip giant's traditional home ground at the high-end of the computer market.
ARM's latest processor core announcement was the Cortex-A15, previously codenamed Eagle. The A15 complies with the ARMv7 instruction architecture but with support for 40-bit virtualization. The next ARM processor to be announced will support 64-bit and could be unveiled as soon as next week, the report said.
One possibility - which would be very reminiscent of Intel's marketing style - would be if the basic Cortex-A15 design already supports 64-bit processing and ARM has quietly kept that detail back from the original announcement to give it more publicity. As ARM is licensor of IP it might be possible to allow chip partners to choose whether to opt for full 64-bit processing or opt for 32-bit depending on application and as they are designing their implementation of the chip.
hes dreaming. he thinks were all going to take step back and accept pc's with single core ghz chips? nobody hates intel THAT much.
hes dreaming. he thinks were all going to take step back and accept pc's with single core ghz chips? nobody hates intel THAT much.
Such speeds are all you need for phones, nettops, entry level notebooks and even some desktops. Eventually, low power CPUs will reach the point where they are all you need in a desktop. And they are starting to be used in servers since you can use so many of them at once in the same box, unlike the more power hungry x86 processors.
no, have you ever tried to watch video on an ipad???? even web surfing is pokey with current arm chips, and thats with fewer pages open that youd have on a desktop. arm will never put intel out of business
The A4 sips power. In fact, power consumption is probably the reason Apple hasn't stepped up performance much from the iPhone. In order to get 10 hours of battery life, the entire iPad (including display) has to pull less than 2.5 Watts on average.
no, have you ever tried to watch video on an ipad???? even web surfing is pokey with current arm chips, and thats with fewer pages open that youd have on a desktop. arm will never put intel out of business
hes dreaming. he thinks were all going to take step back and accept pc's with single core ghz chips? nobody hates intel THAT much.
ARM's big achilles heel is all the discrepancies between implementations from different manufacturers. Just look at any Android app's comments and see how many "don't work on phone so-and-so". The people who use servers will insist on application stability and compatibility. How is ARM going to deliver cross platform compatibility, if they have no control over the final chip implementation?
Maybe they will go back to making the CPUs themselves (or at least subcontracting the manufacture to their exact specs).
That's how ARM started. Acorn computer were a producer of desktop PCs, and they built their own CPU and OS - later spinning off ARM (Acorn RISC machines; later advanced RISC machines). The early CPUs were sold by ARM and built by VLSI. Things changed about 10 years ago, when ARM decided to start licensing their IP to anyone who wanted it - as a result, there are a profusion of proprietary ARM based CPUs all with different properties and proprietary options (enhanced caches, better I/O, AES or RSA acceleration, vector extensions, etc.)
lets not forget that intel is also aiming for arms markets. a year from now intel will have their own low powered cell phone/tablet chips and you can bet that will seriously erode arms markets.
intel itself has already said that in the next few years it will be able to outperform arm at lower power, someone can repost the powerpoint..
http://www.anandtech.com/show/3905/...-headed-for-smartphones-notebooks-and-servers
The Cortex A15 will be available in quad core/2.5 GHz speeds.
I think the consumer computer market is going to diverge. The rise of netbooks and devices like tablets suggests that a good proportion of the PC using population is just using their computer for basically the web, messaging, and other relatively trivial tasks (from a computational standpoint). That market can and likely will move to low power and more convenient devices like tablets and all-in-one PCs, and ARM chips could thrive there.
ARM could well rise into a more important position - not because PCs are going to all be using ARM designed processors, but because people will be able to do everything they use a PC for now on another cheaper, simpler device. There are so many people out there whose computer is basically a Facebook terminal.
Ill take that bet. Where are all those Silverthorne MIDs? Those moorestown tablets and phones? If they are selling at all, they sure aint hurting ARM. Those markets are growing so fast anyway, that even if intel captures some of it, it wont make a dent in ARMs revenue or position. ARM has very little to be afraid of.
THEY WONT BE OUT UNTIL 2011: http://www.maximumpc.com/article/news/intel_prepping_moorestown_cpu_2011_launch
They've said that a few times now. Maybe one they day they will, but that alone wont guarantee intel a lot of success. Too many other issues, not the least of which as pointed out by Hauser is the business model. ARMs approach is just infinitely more flexible, and for socs thats worth more than performance or performance/W. Remember how intels attempt to mimic that somewhat by allowing third parties to integrate atom on TSMC failed miserably?
Another problem for intel is the margins. ARM itself makes a few cent per soc (IIRC 2 cents on average). Its licensees make more, but that cost covers most of the core development and is shared by billions of chips each year. Intel's business model is build on 1000x higher gross margins to pay their R&D. They just dont have that volume.
But even if intel could produce and sell atoms for a similar cost/price as comparable arm chips, how are they going to prevent companies from adopting those $10 chips, rather than $100+ celerons and pentiums for lowend machines? From there it trickles up. We've seen it with atom, it was meant for MIDs and ultimately phones, but it ended up replacing much of intels low end. And that was without direct competition and therefore with all kinds of artificial barriers to prevent just that (screen size limitations, no HD decode, no HDMI, no DVI etc etc).
I just dont see how intel can win this in the long run. Their best hope is probably our dependency on windows, but I doubt that will be enough.
The thing that makes me curious about ARM's prospects in the high-performance space is that this space use to be dominated by higher performance RISC architecture before x86 invaded that space...if the highest performing RISC architectures of the time couldn't prevent domination by the lowliest of CISC architectures at that time then it just seems like an even taller order today for a traditionally low-performance RISC architecture like ARM to somehow step up its game and take on x86 in the performance space.
ARM's big achilles heel is all the discrepancies between implementations from different manufacturers. Just look at any Android app's comments and see how many "don't work on phone so-and-so". The people who use servers will insist on application stability and compatibility. How is ARM going to deliver cross platform compatibility, if they have no control over the final chip implementation?