Ultra-low voltage, or ULV, refers to a range of processors that are more powerful than Intel’s netbook-centric Atom while retaining excellent battery life and allowing for slim figures. (I’m typing on an ULV laptop now, an Asus UL80vt.)

These thin-and-light ULV laptops were pricey when Intel introduced them a couple years ago, and they quickly earned niche status instead of mainstream success. Still, they offer what a lot of people are looking for in a computer — moderate performance and strong battery life in a lightweight frame — and pricing has come down. The company has already launched low-voltage versions of its Core i3, i5 and i7 processors

So it makes sense for Intel to give ULV a bigger role in its lineup. Whereas the the power draw for Intel’s chips previously centered around 35 watts, the company plans to set the center point around 10 or 15 watts, with the goal of making 10-hour battery life a reality for most machines.

On a recent trip to Best Buy, I was surprised by how chunky most laptops look, even compared to my 18-month-old machine. If Intel and PC makers can deliver lots of ultra-thin ULV laptops in the coveted $600 price range, the dreary old laptop could start to look exciting once again.

Seems to me that there are several factors that make a Mac move to ARM plausible, or least very far from unthinkable…

1) We know for a fact that Apple likes to be as deeply involved in the technology inside its devices as possible–that’s why it uses custom processors based on ARM’s technology inside the iPhone and iPad. Intel’s upcoming chips based on “3D” technology sound neat, but they’ll always be Intel chips, not Apple chips.

2) By building its own custom ARM-based chips, it’s at least possible that Apple could make a great leap forward in laptop battery life compared to the rest of the industry. (The iPad already has strikingly better battery performance than other tablets based on non-custom ARM processors.)

3) OS X already runs magnificently on ARM chips–in the form of iOS, which is a variant of OS X as it runs on Macs, not an unrelated operating system. Bringing full-blooded OS X to ARM wouldn’t be an entirely new challenge.

4) Apple has already said it plans to make future Macs more like the iPad. Wouldn’t basing them on the same chips help?

5) Mightn’t the company prefer to build software for one chip platform (ARM) rather than two (ARM and Intel)?

There’s a sixth reason why ARM Macs feel like a real possibility to me–and that’s history. Especially the history of an uncannily similar rumor from a little under six years ago.

On May 23th, 2005, the Wall Street Journal reported that Apple was considering moving from IBM’s PowerPC chips to Intel ones.  The experts quickly chimed in, and said that this rumor was…well, wrong. Or at least almost certainly wrong.

A couple of weeks later, both the Journal and Cnet News.com reported as fact that Apple was about to announce at its WWDC event that it was moving to Intel. This time, with two respected news sources reporting that it was definitively going to happen, the response from multiple experts was still that the idea was far-fetched.

Even Daring Fireball’s John Gruber, who has a better track record of channeling Apple’s thought process than any outsider, couldn’t figure out how to make it make sense.

And then the far-fetched, unlikely, unthinkable scenario turned out to be true:

In other words, rational analysis by knowledgable people turned out to be a lousy way of determining whether Apple was going to bet the Mac platform on a huge transition to a new architecture. Actually, rational analysis by knowledgable people almost always seems to be a lousy way of predicting what Apple is going to do. I’m keeping that in mind as I read about the ARM scuttlebutt, and think about it.

Footnote: After the Apple-Intel news became official, some analysts still thought it made no sense;:

While we can see why moving to a dual architecture approach may bring some benefits, a wholesale move away from the IBM chips would be extremely foolish. Intel is not the ‘de-facto leader in processor design’ that it was a few years ago; in the recent past, Intel has been out-innovated by both AMD (with a better approach to 64-bit computing) and IBM (with a better long-term strategy around multicore chips.

You be the judge of whether Apple knew what it was doing in 2005–and might know what it’s doing in 2011 if it transitions Macs to ARM processors.

Steve Ballmer’s CES keynote isn’t until 6:30pm PT tonight–I’ll be liveblogging it–but Microsoft already made news today at an afternoon press event by confirming the Wall Street Journal’s report that it’s working on a version of Windows that will run on the ARM chips widely used in phones, tablets, set-top boxes and other computing devices that aren’t PCs, as well as competitive x86 system-on-a-chip designs from Intel and AMD. Windows honcho Steven Sinofksy did some demos of Windows (and Office, and IE) running on test boards powered by these processors, and said that the system requirements of phones and the system requirements of PCs are starting to converge, and that his demos were of “the next generation of Windows,” which he refused to call Windows 8. He also showed a new version of Microsoft’s Surface table build by Samsung and based on all-new technology.

And that’s about all he did–he cheerfully announced that he wasn’t talking about the user interface of the new Windows or when it might ship. More thoughts later…

I was startled by the news–until I thought it over, whereupon it didn’t seem so surprising any more. For decades, x86 processors (mostly from Intel and AMD) have been inside most computers that mattered, and so the fact that Windows ran on them was a virtue. (In fact, when Microsoft ported Windows NT to other CPUs in the 1990s–DEC’s Alpha and MIPS–the new versions turned out to be irrelevant, and so the company pulled the plug.)

But what happens if tablets and other new-wave computing devices become serious rivals to traditional PCs? x86 as it stands today isn’t especially well-suited to tablets, since it wasn’t designed from the ground up for energy efficiency and small form factors. (That was supposedly one reason why HP pretty much lost interest in its own Windows tablet and bought Palm’s Web OS.)

And even if Intel and AMD start to build x86 processors designed with new types of devices in mind–and they will–being x86-only lashes Windows to those companies’ product road maps. It doesn’t let Microsoft control its own destiny.

The one aspect about this news which is (very) surprising is one tidbit in Clark and Wingfield’s story: Even though Microsoft will supposedly show the ARM version of Windows, the new edition reportedly won’t be ready for a couple of years. That’s an eternity–and it means that Windows might not truly be competitive as a tablet operating system until some time in late 2013. (And even a Windows that runs on ARM chips won’t be competitive unless Microsoft radically reworks its interface and third-party developers build tablet-friendly applications.)

If all this news is accurate, it means Microsoft has been awfully slow to figure out where the computing world is going and start marching in that direction. A truly prescient Microsoft would have figured out some of this stuff back in 2006 or thereabouts–or at least by 2007, when the iPhone came out–and would be shipping Windows for ARM right now.

Speaking of the iPhone, Apple pulled off the evolution of OS X almost flawlessly over the past few years, starting in 2005, when it moved OS X from PowerPC chips to Intel ones. In 2007, it released an ARM version of OS X with an all-new interface, and everything was downsized to fit on the iPhone. In 2010, it took that iPh0ne version (now known as iOS) and created a variant for the iPad. All of this was hard work, but Apple didn’t start talking about any of it until the new versions were nearly ready to go.

It’ll be fascinating to see what happens if Windows begins to move in a similar direction–but Microsoft, in its Microsoftian fashion, starts telling the world about its plans years in advance.

The processor, which Intel calls a “single-chip cloud computer,” is about 20 times more powerful than Intel’s most powerful six and eight core processors that are available on the market today. It also provides that capacity while remaining energy efficient.

It might sound revolutionary, but it is just the evolutionary progression of the many-cores trend that has occurred over the past several years. Intel showed off its ability to design an 80 core chip in 2007, and very little has changed from the end user’s perspective over the past two years.

Increasing the clock speeds of silicon chips was becoming an unsustainable practice due to the diminishing returns in performance that were gained against the increase in costs, energy and heat that was being generated by even a modest uptick in speed. That’s why we have multicore hardware in our PCs today – it’s the only practical way to make them more powerful.

Parallel computing also ended the free ride that programmers had gotten with each succeeding generation of processor becoming more powerful. It’s now time for the average programmer to handle challenges that had only come up in academia.

The promise is applications that can do what was previously impossible on a desktop machine, but companies like Intel need to make that relatively easy to do without requiring too much specialized knowledge. Parallel computing allows many jobs to happen at once, so, for example, a chess program could consider every possibility before making a move.

There are also performance benefits; an operating system could scale in performance when more cores are added, and Web sites will be capable of handling more transactions. MySpace is already using parallel technology from Microsoft to manage distributed transactions across its server farms, for load balancing and failure handling.

But programmers must first understand how to create applications that work well with many core processors before your desktop computer becomes an unbeatable chess player. Democratizing those skills will require education, new languages, extensions to existing languages, specialized tools, and even new operating systems. Intel’s 48 core processor is a catalyst for that work to happen, and its arrival does not mean that the future is now.

The agreement doesn’t end legal action against Intel by government officials, such as the EU’s $1.45 billion fine for abusive business practices (which Intel is appealing) or New York State’s recently-filed lawsuit.

For consumers, the major question about the settlement is pretty simple: Does it increase the likelihood of healthy competition between Intel and AMD, thereby driving greater chip innovation and lower prices so that we get the most PC possible for our money? We’ll see. But it’s fascinating to look at what Intel has agreed to refrain from doing, as reported by Cnet:

• Offering inducements to customers in exchange for their agreement to buy all of their microprocessor needs from Intel, whether on a geographic, market segment, or any other basis

• Offering inducements to customers in exchange for their agreement to limit or delay their purchase of microprocessors from AMD, whether on a geographic, market segment, or any other basis

• Offering inducements to customers in exchange for their agreement to limit their engagement with AMD or their promotion or distribution of products containing AMD microprocessors, whether on a geographic, channel, market segment, or any other basis

• Offering inducements to customers in exchange for their agreement to abstain from or delay their participation in AMD product launches, announcements, advertising, or other promotional activities

• Offering inducements to customers or others to delay or forebear in the development or release of computer systems or platforms containing AMD microprocessors, whether on a geographic, market segment, or any other basis

• Offering inducements to retailers or distributors to limit or delay their purchase or distribution of computer systems or platforms containing AMD microprocessors, whether on a geographic, market segment, or any other basis

• Withholding any benefit or threatening retaliation against anyone for their refusal to enter into a prohibited arrangement such as the ones listed above.

Basically, Intel’s agreeing not to take actions that would shut AMD out of dealing with major PC companies entirely, or hobble it so severely that it might as well be shut out. Sounds good to me. I wanna have the opportunity to choose between PCs based on a variety of processors from multiple companies.

Ultimately, AMD has always fared best when its portfolio of chips has been at its strongest in comparison to Intel’s offerings. Today’s agreement won’t have any immediate effect on its product lineup, of course. But if it increases the chances that a great AMD chip will get a great response from the industry, it would be…great.

Now AMD is making a dramatic bid to simplify branding of its CPUs down to the bare essentials. In fact, rather than emphasizing specific CPUs at all, it will focus on three levels of performance:

PCs with Vision technology are basic machines designed for Web browsing, music listening, and the like. Ones with Vision Premium are potent enough to handle video and audio conversion well, as well as gaming. And Vision Ultimate indicates that a PC is well suited to video recording, audio editing, advanced photo editing, and the like.

Beyond the fancy stickers, there are two simple ideas here: AMD is emphasizing media applications (which makes sense, since video and audio-related performance is the main reason to worry about what chip you get at all) and is giving consumers the classic choice between good, better, and best. (However, it plans to introduce Vision Black, a sort of “bester” designation aimed at gamers and enthusiasts, early next year.)

Intel, meanwhile, is trying to simplify performance comparisons, too–but its menu of choices is broader and more complicated, and it’s not always easy to figure out how everything relates. Which brings up an issue with Vision that’s out of AMD’s control: The most important CPU comparisons are those you make between processors from competing companies, and it isn’t obvious how the three Vision options map to Intel’s chip family.

I’m sure that serious tech enthusiasts will squawk that AMD is dumbing things down too much (and the company does say that it’ll use more traditional, meaty technical facts to market its chips and technology for that crowd). But when I think about how I buy PCs these days, the Vision distinctions would probably do the trick. There was a time when I dithered over whether I needed a CPU with a math coprocessor, and got excited over stuff like MMX extensions. Today, I mostly want a general idea of whether the processor will be potent enough for the tasks I’m likely to throw at it. And once I’ve plunked down my money for a computer, I tend to forget what’s inside.

How much time do you spend thinking about CPUs these days?

Got nominations for other companies you'd like answers from? Let me know--I'm lining up subjects for future installments.]

Fernando Garcia asks:

I have always asked the following question. Why is it that AMD will not step up advertising? A good 70% of the consumer public,still does not know what AMD is. I used to work for Best Buy and on the average day, one out of eight persons I would speak to knew what AMD was. Whenever I asked a customer  about processors automatically they would say Intel.

Pat answers:

Simply taking out more advertising does not guarantee a product’s success. I think the best way to answer that is AMD chooses to focus differently. We first focus on making our customers and their channel partners successful by investing in them, not leveraging off their brandsby sandwiching them between AMD logos. We want to invest in our customers’ success. For those people who are specifically focused on the “processor,” we have very high awareness and market directly to end user groups. These include but are not limited to enthusiasts, gamers, DIYers, Fortune 1000 and government decision makers, etc.

Ron Harris asks:

Does or will AMD offer a chipset for the competition’s Core i7 processors?

Pat answers:

We offer ATI Radeon HD GPUs for the Core i7 but not chipsets.

Justin Abrantes asks:

Why did AMD decided to manufacture quad cores under the name of Athlon? The same way with dual-core Phenoms–will that be more confusing for non-AMD users?

Pat answers:

We did this to simplify, not confuse. AMD Phenom processors generally have more features than AMD Athlon processors. So generally, Phenom>Athlon. X2, X3, and X4 just indicate the CPU core count.

Bill Pytlovany asks:

What’s the status of the new GlobalFoundries’ chip fab plant planned to be built in upstate (Malta) NY?

Pat answers:

I have an inquiry into GF folks, but in the meantime here are the most recent details. Our understanding is that GP plans to break ground this summer on a new 4.2B plant in Malta, NY, with production scheduled for 2012.  It’s too early to specify exactly what products will be manufactured there but we understand it will be a mix of AMD and non-AMD products.  Of course, since AMD and GF are separate companies, we recommend you contact GF directly for further information.

Toby Hudon asks:

Where’s the 4770X2? Any chance of a sub-9″ board for one to fit in those cramped HTPC cases?

Pat answers:

A dual ASIC board utilizing the ATI Radeon HD 4770 chip is a great concept idea. This would be up to our partners to bring to market. I would suggest a single GPU solution for an HTPC to help minimize heat and power and maximize performance per watt.

Toby also asks:

Where’s OpenCL? Nvidia has some kind of super secret beta developer club you can try and beg your way into, but AMD is silent even though they were showing demos on CPUs and Havok on GPUs at shows already. When can we get a compiler and libraries?

Pat answers:

AMD has been and continues to be a leader in adopting and supporting open standards, OpenCL is a very important open standard that we are heavily invested in and expect to have more details available publically shortly

Cliff Forster asks:

AMD has recently reported a gain in market share. To what would you mostly attribute this positive upswing? Social Media strategy paying off? Perhaps the Intel EU press was very damaging to big blue? Recognition of the true dollar for dollar value position in a weak economy? Specific innovations that have improved AMD’s product offering? If you had to pick something specific that is propelling the increase in share, what would you say it is, and how do you continue to ride this positive trend?

Pat answers:

We are in a quiet period and cannot specifically comment on market share, unfortunately. But we can restate that we believe AMD platforms offer maximum value at every price point.

JustCallMeBen asks:

What’s going to happen to Hybrid-CrossFireX? As a gamer caring about the environment (AND my energy bill), I would very much like to use a high-end card like the HD4890 in combination with an Intrgrated Graphics Processor. This way the discrete chip is only used for games while the IGP does the typical desktop usage, thus saving power. Regrettably Hybrid-CrossFireX is currently only available for a limited number of HD3xxx chips. Are there plans to renew the technology?

Pat answers:

AMD has led the market with innovative features like ATI Hybrid CrossFireX technology. Again, we’re in quiet period and can’t comment on specific plans, but innovating platform level features remains an AMD priority.

Mark Davis asks:

In the last 2 years I believe Intel has adopted a ‘Tick-Tock’ engineering schedule, in one time frame putting out a new product, then in another a refresh and then back to a new product. What kind of engineering schedule does AMD follow? And are we most likely to see a new micro-architecture for Bulldozer?

Pat answers:

AMD and Intel have similar engineering schedules; it’s natural given the competitive environment. AMD is near fully ramped on 45nm process, roughly around the same time as Intel got fully ramped. So it is a misconception to think Intel is years ahead of AMD in terms of manufacturing technologies. Our technology partner GlobalFoundries has indicated that it will shortly breaking ground on a state-of-the-art 20nm fab facility in New York State. When this facility is finished some time in 2012, it is expected to be the most technologically advanced manufacturing facility in the world.

Mark also asks:

Why does it seem like ATI’s game developer relations are so lackluster compared to Nvidia’s? For example on most game boxes I see n Nvidia The Way It’s Meant to Be Played ad versus an AMD ad. Are there any plans to change this?

Pat answers:

Let’s avoid confusing Developer Relations with marketing. TWIMTBP is a marketing/co-marketing program that negotiates game asset usage and NV logo placement; AMD has a program like this as well but it has not as big a focus for AMD as TWIMTBP seems to be for Nvidia. Check out Battleforge for example. As for Dev Relations, this is an engineering relationship between the dev and our engineers to optimize new games to help achieve the best customer experience. We are highly invested in this program with a very wide breadth of game and application developers.

Stilgar asks:

I would like to know what processors AMD is working on that are not targeted at desktops/laptops? Do they have a response for the Atom? What about cell phone CPUs/GPUs?

Pat answers:

We are constantly analyzing new potential markets. For example, along with our technology partner HP, AMD helped create the “Ultrathin” mobile market with the introduction of the HP Pavilion dv2. We have no plans to specifically target the mobile phone market. Our roadmaps target server, notebook and desktop platforms are available in myriad form factors. And of course, the top two selling game consoles in the world are powered by AMD graphics processors, so the console market is important to AMD.

Wayne Smith asks:

Recently I had to replace my processor (Athlon dual core 64 5200+) which was rated at 2.6-GHhz with 1024MB L2 cache. The replacement processor (same 5200+) is rated at 2.7-GHz with 512 L2 Cache. Is there any difference in performance between the two processors?

Pat answers:

The AMD Athlon X2 5200 processor comes with 2MB total cache (1MB per core) or 1MB total cache (512KB per core). There is no meaningful performance difference.

George asks:

I am a longtime AMD fan and I have not bought a computer if it did not have an AMD CPU in it. Of course now I build my own still using AMD and I just was thinking that when you went from single core to dual-core and also again from 32-bit to 64-bit, that the experts thought that the performance might nearly double but that was never the case. It seems to me that you come out with CPUs that have the bare minimum number of pins to support the technology while Intel seems to be a hundred or more pin counts ahead of you. With that said, how does the pin count affect the efficiency of the CPU? And have you considered significantly adding the number of pins, since you did not when you converted from single core to multi-core?

Pat answers:

Pincount is related to the system architecture the processor supports. With AMD processors, the evolution from single channel memory support with Socket 754 to dual channel 939 support was the major pincount increment, Subsequently, the migration from DDR to DDR2 and now DDR3 are associated with changes in pin assignments as our processors feature an integrated memory controller. These are the main drivers of changes in pin count and socket configuration over time. Pin count does not have a direct bearing on CPU frequency, but AMD continues to refine our socket and our architecture to deliver optimal performance.

]]> http://technologizer.com/2009/07/07/your-questions-amds-answers/feed/ 2 patrickmoorhead Technologizer;s Q&A http://technologizer.com/2008/10/07/amd-splits-in-two/ http://technologizer.com/2008/10/07/amd-splits-in-two/#comments Tue, 07 Oct 2008 19:23:30 +0000 Harry McCracken http://technologizer.wordpress.com/?p=2741 ]]> It’s been a possibility for a while, and now it’s a reality: AMD, the perennial number-two CPU company to Intel and one of the few chip companies that both designs and manufactures processors, plans to break itself up. The company behind Phenom, Athlon, Opteron, and other CPUs will become two companies: one that designs chips, and one that makes them. The design company will end up partially owned by Mubadala, a company which is owned by Abu Dhabi; ATIC, another company owned by Abu Dhabi, will own the majority of the manuacturing company. Both of those Middle Eastern investments will provide an infusion of cash which is designed to help AMD with its next-generation chips and therefore its overall competitiveness with Intel.

Emotionally, the move may be a big deal for AMD, which has spent decades taking on Intel by, essentially, trying to be Intel. But nearly everyone else involved in the designing and building of processors has decided that financially, it makes sense to separate the building part–which involves massive, massively expensive plants–from the designing.

I’m neither an economist nor an expert on chip manufacturing, so I can’t judge the deal on its merits. But if it helps the two new companies produce more advanced chips more quickly, it’s a good thing for consumers. And, of course, a good thing for AMD, which has struggled to stay even vaguely competitive with the products from its much larger competitor in recent years. (The golden age of the Intel-AMD wars were back around the turn of the century, when AMD rolled out the excellent original Athlon CPU, giving every PC user a reason to consider an AMD-powered computer–and giving Intel a scare that ensured it wouldn’t spend the next few years resting on its technological laurels.)

The chip wars matter to most consumers only because they’re a driver of healthy competition that results in faster, cheaper CPUs that power faster, cheaper computers and other devices. For that reason, I’m happiest when AMD is at its most competitive versus Intel–and hope that this corporate breakup makes as much sense as AMD thinks it will.

]]> http://technologizer.com/2008/10/07/amd-splits-in-two/feed/ 0 Harry McCracken amdsplit