Convergence and the Changing PC Landscape

It's too cliché to proclaim netbooks are dead. Perhaps the appropriate phrase is netbooks are no longer interesting to write about, but they do have a roadmap going forward. For years we heard about convergence in the PC and consumer electronics space. Our TVs, set top boxes and video players were all supposed to get more intelligent. Last year we saw the first real fruits of those efforts with the introduction of Google TV and devices like the Boxee Box. Convergence has finally reached mainstream, but the process isn't over yet.

The smartphone revolution is the beginning of a much larger convergence. A melding of computing devices, convergence between the smartphone and tablet, or the tablet and notebook PC. As is typically the case of any convergence process, this one doesn't have a clear end nor does it have a clear roadmap. Along the way there will be many attempts and likely more failures than successes until we meet the handful of devices and strategies that really make sense.

The smartphone will become even more PC-like and the tablet will become even more notebook-like. But where does that leave PCs? They too must evolve and play their role in this process of convergence, after all it will be PC technologies that ultimately drive the convergence (from the example earlier, is it any surprise that both Google TV and Boxee Box feature very PC-like processors running Linux based OSes?).

How does the PC evolve? Part of it is a change in software. Traditional desktop OSes won't go away, but they must incorporate the advantages and innovations brought by the players in the smartphone/tablet markets. Apple is at the forefront of much of this revolution and the next version of Mac OS X already starts to look more like iOS with its app store, app launcher and iOS-like full screen modes. Microsoft is rumored to be preparing a very tablet friendly UI that will layer upon Windows 8, which itself will span everything from tablets (and smartphones?) to desktops.

As we've learned in the past, software enables hardware and hardware enables software. The PC's changing role in the future also requires some new thought about hardware design and what sort of decisions microprocessor manufacturers are going to make going forward. We've already seen hints of this from both AMD and Intel. The two companies no longer make discrete CPUs but rather complete SoCs, similar to what you'd find in a smartphone just on a much larger scale. Like I said before, the PC will adopt the learnings of the smartphone and tablet industries as it evolves.

Today Intel is announcing the first step in that evolution, an announcement that we actually first heard about from another company a year ago.

When Apple introduced the 2010 MacBook Air, Steve Jobs called it a preview of the future of the MacBook lineup. The subsequent MacBook Pro release looked pretty traditional so the messaging may have been a bit premature. I believe what Jobs was referring to was the move toward thinner, lighter and SSD based notebooks across the board. Intel's announcement today puts that future on a roadmap, and the device is called the Ultrabook.

Meet the Ultrabook

The Ultrabook is a multi-year evolution from Intel's perspective. It's going to begin as a pilot program with some Sandy Bridge systems this year, it'll ramp heavily next year with Ivy Bridge and be mainstream by the time Haswell arrives in 2013.

What is an Ultrabook? It's basically a thin and light notebook that uses solid state storage in some form, has some additional security features and is available at mainstream price points. This isn't a new platform, there's no Centrino-like certification process, but Intel has trademarked the name so you won't see things that aren't Ultrabooks being called Ultrabooks (unfortunately this also likely means that you won't see any AMD notebooks being labeled as such either).

The first requirement of an Ultrabook is that it's ultra-thin. In Intel's eyes this means less than 0.8" (20.32mm) which is thinner than anything Apple offers in the MacBook Pro line (0.95").

The second requirement is that the system needs to be ultra responsive, either through the use of an SSD or SSD caching. Intel was quick to point out that an Ultrabook doesn't have to use an Intel SSD, it just needs to have SSD-like response time.

Here's where the requirements start getting vague. Intel asserted that Ultrabooks need to be secure. Today that security comes by way of Sandy Bridge, which offers Intel's Identity Protection Technology (IPT) - basically a unique hardware token embedded in the SNB CPU. You tell an application that your computer is secure, and going forward it uses the presence of that unique token as a form of authentication.

Ivy Bridge will add some additional security features (on-package digital random number generator, and higher level execution protection bit) and I'm sure Haswell will go even further. Remember Intel's acquisition of McAfee? I suspect that's going to be a part of this security strategy.

The final requirement is all Ultrabooks must be sold at mainstream price points, which Intel calls sub-$1000.

You'll notice that I haven't mentioned Atom. That's on purpose as Atom will not be driving any of these Ultrabooks. Intel has separate plans for Atom, for use in smartphones and fanless designs which I'll get to in a moment. But Ultrabooks are the future of mainstream notebooks as far as Intel is concerned. A Sandy/Ivy/Haswell based notebook that's less than 0.8-inches thick, has some form of an SSD and is sold for less than $1199 is a-ok by me.

One of Intel's closest partners, ASUS, already introduced the first Ultrabook at Computex: the new UX21. Availability of ASUS' UX Series Ultrabook is slated for Q4 of this year:

Intel estimates that by the end of next year 40% of consumer notebooks will be Ultrabooks. Given the desirable set of features and reasonable price point, I can see that happening.

Always On, Always Connected

Intel views the Ultrabook as a new category of mobile devices, however everything I've described thus far sounds a lot like a thin and light notebook with an SSD inside it. There's a software component to all of this that Intel is promising, starting as early as Sandy Bridge.

Intel wants to bring the instant-on capability of tablets to Ultrabooks. Apple already did some of this with the new MacBook Air. Suspend to NAND allows for a reliable method to quickly hibernate and resume. ASUS is already promising a 2 second resume from sleep time on its new UX Series.

Intel also wants to bring the always on, always connected experience to Ultrabooks. 3G tablets and smartphones currently enjoy this but notebooks have lacked it. When you wake your notebook up from sleep you usually have to wait to download all new emails, receive all new twitter updates, etc... Through a software layer Intel isn't ready to talk about yet, Ultrabooks will be able to pull this data from the cloud while the machine is otherwise asleep. That's the functionality with the first generation of Ultrabooks; the second generation will move to a smartphone-like push model where servers push this data to your Ultrabook. Again, details on how this will all work are basically nonexistant at present but the goal is to take some features from the smartphone/tablet space and add them to a much sleeker, sexier notebook. Hence the name Ultrabook.

A Changing Thermal Target: Discussing Haswell

Pat Gelsinger once taught me that a single microprocessor architecture can efficiently target an order of magnitude of TDPs. It's not that you can't scale above or below that range, but at that point it becomes more efficient to use a different microarchitecture.

Let's take Sandy Bridge for example. Current desktop variants of the chip exist at 65W and 95W TDPs and later this year we will see Sandy Bridge E with a 130W TDP. If we pick 130W as the upper bound for the architecture, it should be able to efficiently scale down as low as 13W - or one order of magnitude. Looking at mobile SNB processors, it does.

Intel's ultra low voltage SNB carries a 17W TDP, while mainstream mobile SNB chips are in the 35W (dual core) to 45W (quad core) range. These TDPs all include processor graphics. With Sandy Bridge, Intel has an architecture that spans from 17W all the way up to 130W. I wouldn't be too surprised if we eventually saw some ~13W SNB parts for really low power applications in the future.

At the end of Q1 of next year Intel will introduce Ivy Bridge, its first 22nm microprocessor. I fully expect Ivy Bridge to target relatively similar TDPs as Sandy Bridge, however the initial launch will be confined to TDPs less than or equal to 95W (much like SNB was).

With Lynnfield Intel made it very clear that it's possible to get high-end desktop class performance out of a 95W part. While the 130W chips were still faster, the majority of enthusiast users would get by just fine with Lynnfield. The move to Sandy Bridge highlighted Intel's move away from 130W TDPs for high-end desktop processors and down towards 95W. As I noticed in my transition to a mobile quad-core SNB notebook as my primary workstation, I believe this generation of 130W CPUs will target a smaller portion of enthusiast users than the previous generation. The trend is definitely downward, towards lower TDPs.

Haswell is where Intel's architectures take a dramatic turn. Ivy Bridge is a derivative of the Sandy Bridge architecture, which of course was designed for that 13 - 130W range. Haswell however is a brand new architecture. It'll likely look similar to Sandy and Ivy but its target TDPs will be shifted down. In mobile, Haswell designs will be set at 10 - 20W. That's not the lower bound of the design, just the target for mobile. What does that do to the rest of the scale? Intel presented this slide at its analyst day earlier this month:

In Sandy Bridge, mobile occupied the 35 - 45W range - roughly the bottom third of the architecture's target. Around Haswell two things happen: the mobile design drops and the Atom design target moves upward.

Atom will service a new expanded range from ~800mW to 8W, leaving Haswell to address the ~10W and above market. Multiply that number by 10 and we have our upper bound of 100W - which isn't much different from the 95W we see today for high-end SNB SKUs. That being said, I do believe we'll see a lot more focus around 65W in the desktop.

Where Does This Leave Atom?

When Haswell hits in 2013, Intel will also update the Atom line to introduce the recently announced Silvermont architecture. As I speculated publicly before, based on Intel's hint to me five years ago that Atom would remain in-order for five years, I'm expecting Silvermont to be Intel's first Out-of-Order Atom.

When Atom was introduced, its microprocessor architecture was unlike anything we'd seen from Intel in years. It looked more like a modern take on the Pentium rather than a P6/Conroe derivative. That was nearly five years ago, so what do we expect Silvermont to look like? I'd say it might look a lot like a modern, ultra low power take on Conroe. Perhaps a limited Out-of-Order execution core, only a 3-wide front end and a number of other limitations to keep its power in check.

In the sub-10W range Intel is going to use Atom specifically for fanless designs - the next generation of netbooks as well as tablets. Cedar Trail is Intel's 32nm Atom platform which we've talked about already, which will be used in netbooks such as ASUS' newly announced EeePC X101:

Medfield is the ultra mobile version of Atom, destined for handsets and tablets, and at the show we should finally see a Honeycomb (Android 3.0) tablet running the SoC. Unfortunately with NVIDIA Tegra 2 based tablets moving to Android 3.1 already, and Medfield designs not due out until later this year I'm wondering if that may be too little too late.

Expect more coverage on Intel's announcements at the show as we come by them.