So, Intel had basically the same CPU from January 2010 to today. That's over three years of no competition to power efficient, but pretty weak ARM (ARMH)-based tablet computers. For Intel, there is no shortcut to these incredibly complex devices and manufacturing processes. The Tick-Tock cadence is shrink, new architecture, shrink, new architecture, wash and repeat. For the past three years, Intel has been a slave to the strategy that will, ultimately, come to obsolete the ARM-based tablet toys.
It seems to us that he's not entirely convincing, or even logically consistent though. In extolling the benefits of these new frugal processors from Intel, he stresses the fact that they reduce idle power to 5% of the preceding CPU. Since PC CPU chips run idle most of the time:
Running video, calculating big spreadsheets, game playing, etc. tax the processor more, but in the end your PC CPU chip is idling the vast majority of the time, so the 95% reduction in idle power is a big deal. [Russ Fisher]
And he concludes that:
As the above begins to be understood later this year, we might expect a large spike in WinTel PC sales. When consumers see that they can truly "have it all", we might see some backlash against those who took computer level money for a toy masquerading as a computer. The Intel multi-year misery in the PC market will be over. Right after that the Intel crosshairs move to the smartphone market.
If PC CPU's run idle most of the time, we fail to see how he can call ARM-based tablets as toys masquerading as computers. And indeed, one ask, what can't a tablet do that a PC can, apart from really taxing applications that few users outside business seem to be really wanting. Tablets can even play games pretty well, which is something numerous PCs actually struggle with (which is why there is a special category of PCs called gaming PCs).
To us, it seems that the overwhelming majority of users seem perfectly happy with the benefits of their ARM based toys in terms of portability, speed (instant on, instantly loading applications, something PCs still struggle with), better screens (pixel packed IPS panels on the high-end tablets) and all-day battery life.
Even in previously PC dominated settings, like business, healthcare and education, tablets are really moving in strongly with a slew of innovative applications taking advantage of the inherent advantages of tablets we just explained. So we see no need to change our previous stand and we don't expect a revival of the PC, as we earlier argued (here, here and here).
However, the Haswell-based processor does enable Intel-based devices to finally compete with ARM-based devices in terms of energy frugality. To us, that means that Intel-based phones, tablets and hybrids will get a boost. Apart from Intel itself, Microsoft (MSFT) could also benefit here.
The advantage of the Windows 8 based tablets and hybrids was supposedly that it can run legacy software, but the disadvantage was that there was an awkward trade-off between power and energy and cooling needs. The user could either buy a rather underpowered Atom CPU based tablet/hybrid that was as sleek and energy efficient as most of the competition, or a much less energy efficient iCore based tablet/hybrid that is thick, heavy, doesn't last all day but is fully powered in terms of processor needs.
With the new Haswell processors, that trade-off has become distinctly reduced, perhaps it even disappears completely. Of course, one could buy the ARM-based Windows tablet/hybrid, but we really failed to see the point of that, as apart from Works, it doesn't run legacy Windows software and in Q1 of this year, only 200,000 WindowsRT devices were sold.
Another Intel processor disadvantage is simply price:
Lenovo this week announced an Intel-powered version of the 11.6-inch convertible/hybrid laptop-cum-tablet Yoga. The Intel Yoga is at least $240 USD more than the Tegra 3-based Yoga retails for around $560 USD
And that difference rapidly increases if you want a faster Intel processor, a Yoga powered by a high-end Core i7 processor from Intel cost $1349.99. Although most people still apparently prefer the more expensive (Intel powered) Yogas, but that's merely a function of the weakness of the Windows RT operating system.
But we have to say, the jury is still out whether Windows tablets/hybrids will really take off, the arrival of more frugal Intel processors is helpful, but whether it is enough remains very much to be seen.
Progress in ARM land
But we shouldn't be blind to the fact that progress in chip performance isn't limited to Intel only. While Intel processors are getting more frugal, ARM-based processors are getting more powerful. NVIDIA (NVDA) for instance has just rolled out the fourth generation of its Tegra chip, one with 72 GPU cores and a separate LTE modem chip, matching current offers from Qualcomm (QCOM) Snapdragon 800.
By the fourth quarter, the Tegra 4i will be just half the size of that, with significant progress following next year with the arrival of 64-bit Tegra's. And the likes of Qualcomm and Samsung (SSNLF.PK) are not sitting still either, nor ARM itself.
So while Intel chips are getting more frugal, ARM-based chips are getting more powerful. One could argue that since few people having the latest ARM-based tablets and hybrids complain about a lack of power, Intel's progress in frugality matters more and that's probably right.
However, what's also probably right is that most consumers do not care a great deal about the type of processor that is powering their gadgets, they care about the experience and switch only if that experience has awkward flaws and trade-offs. The only awkward flaw that ARM-based tablets and hybrids are likely to have is that they don't run a host of legacy Windows software.
This clearly matters for business, even though tablets are taking on whole new swaths of functions in business. But as long as good alternatives exist for consumers, this matters a great deal less to them. Since we believe this to be the case, we cannot really see any big shift from ARM-based tablets and hybrids towards Intel/Windows-based ones, unless OEMs make that shift for consumers.
We think there might be some of that, but let's not forget that the ARM/Android-based tablets and hybrids are significantly cheaper as well, and Google (GOOG) is rapidly increasing the Android user experience.
Microsoft faces considerable additional hurdles in the competition with Android and Apple-based tablets:
- Momentum, Apple (AAPL) and Android have already established quite a dominant position
- Lack of applications, compared to the Apple and Android ecosystems
- Awkward trade-offs in Windows 8, which serves both desktops and tablets.
The first two are fairly difficult to deal with, apart from offering huge incentives to application developers to overcome the disadvantage of having a much smaller market share. And the growth in the number of apps seems to be slowing distinctly:
Last June, the Windows Phone Store saw tremendous growth, doubling the number of apps in a six-month period to total 100,000. In the past 11 months, however, less than 45,000 new applications were added to the marketplace. [Yahoo]
Still, one is inclined to say 145,000 apps is a significant number, enough to have a viable platform. However, the relative number (compared to Android and iOS) is what counts most for new customers, and here the disadvantage is still notable and not getting better.
The chicken and egg problem of coming up short in the number of apps, which is a problem attracting new customers, which is a disincentive for app developers has also manifested itself lately, after the initial enthusiasm has worn out. The third disadvantage is not really relevant for tablets, Apple CEO Tim Cook compared it to trying to merge a refrigerator and a toaster.
But since we think that most of what people can do on a desktop they can (or soon will be able to) do on a tablet, they should center on that. Clearly the strategy is directed at efforts to capture parts of the Android and Apple tablet markets, but enable executives to do some serious work on their Surface Pro as well.
Like many observers, we think that re-introducing the start button (which rumor has they will, but not the start menu) would go a long way in achieving this. Other possible improvements are the ability to boot in desktop (also rumored to be in the Blue update), eliminating the reliance on keyboard shortcuts and allowing apps to run in a window
We think another useful improvement would be to enable legacy software to reappear as tiles on the Surface screen that can be activated by mouse and touch alike. Suddenly, Microsoft's ecosystem would have many more "apps" if they did.
How difficult can that be? We're no software specialists, so we cannot give you an informed opinion about that, but we can't imagine it's beyond reasonable possibility. If they can do that, they suddenly have a unified operating system for the desktop and mobile world.
While Intel's new frugal processors are mainly an opportunity for Intel to crack the tablet and smartphone markets, there is something to be said for the proposition that this will increase Microsoft's chances in that space as well (mainly tablets/hybrids, as legacy Windows software isn't terribly relevant for smartphones).
The new Haswell processors enable Windows 8 based tablets/hybrids that are as energy efficient as ARM-based stuff and the Windows 8 based tablets/hybrids could become a lot sleeker and more attractive as a consequence, due to the fact that these run Windows legacy software.
However, the problems with Windows 8, the base and momentum behind the ARM/Android ecosystem and the fact that ARM-based processors are becoming ever more powerful are really important impediments for any Microsoft assault on the tablet/hybrid market. We remain quite skeptical.