Intel announced that the Clover Trail series of processors will be "Windows only", lacking the disclosed information necessary to support Linux. But Clover Trail is already a dead end for other technical and business reasons.
Clover Trail is said to include power-management that will make the Atom run longer under Windows. It had better, since Atom currently provides about 1/4 of the power efficiency of the ARM processors that run IOS and Android devces. The details of Clover Trail's power management won't be disclosed to Linux developers. Power management isn't magic, though - there is no great secret about shutting down hardware that isn't being used. Other CPU manufacturers, and Intel itself, will provide similar power management to Linux on later chips.
Why has Atom lagged so far behind ARM? Simply because ARM requires fewer transistors to do the same job. Atom and most of Intel's line are based on the ia32 architecture. ia32 dates back to the 1970's and is the last bastion of CISC,
Complex Instruction Set Computing. ARM and all later architectures are based on RISC,
Reduced Instruction Set Computing, which provides very simple instructions that run fast. RISC chips allow the language compilers to perform complex tasks by combining instructions, rather than by selecting a single complex instruction that's "perfect" for the task. As it happens, compilers are more likely to get optimal performance with a number of RISC instructions than with a few big instructions that are over-generalized or don't do exactly what the compiler requires. RISC instructions are much more likely to run in a single processor cycle than complex ones. So, ARM ends up being several times more efficient than Intel.
ARM adds some complexity by supporting three additional instruction sets on some CPUs: Two 16-bit versions of their 32-bit instructions, and additional instructions that are optimized for running a Java virtual machine. Even with this added complexity, ARM provides several times the power efficiency of Atom.
So, we start with the fact that Atom isn't really the right architecture for portable devices with limited power budgets. Intel has tried to address this by building a hidden core within the chip that actually runs RISC instructions, while providing the CISC instruction set that ia32 programs like Microsoft Windows expect. But this doesn't approach ARM's power efficiency.
Clover Trail is further handicapped by the presence of
PowerVR graphics, also known as
SGX and
Intel GMA 500. Intel doesn't own the PowerVR architecture, it's just licensed it for Clover Trail and a few previous processors. Imagination Technologies, owner of PowerVR, doesn't play well with Linux developers. Intel previously used PowerVR in their
Poulsbo architecture, a disaster best explained by Phoronix's recent article
Years Later, Intel Poulsbo Remains a Bloody Mess.
Clover Trail's target is a future
Windows 8 Tablet. Given the lack of Linux support and the remote probability that Apple would be interested, that's the
only platform available to the CPU.
If you expect the Windows tablet to do as well as the Windows 8 smartphones recently released by Nokia and others, you probably aren't far from wrong. Clover Trail, built with partner Microsoft, might be Intel's biggest loser since Itanium, built with partner HP.
Why does Intel enter into expensive partnerships to build these losers? Itanium, had it succeeded, would have offered Intel an exit from the ia32 architecture. Intel sorely needs that exit, but hasn't been able to transition its
customers off of the ia32 instruction set. ia32 has been "patched" to keep it competitive for close to 40 years now, most recently with x86-64, an exact copy of the 64-bit extension that AMD created for its own processors. But ia32 is horribly obsolete and ultimately a dead-end for Intel, so their choice to enter into Itanium is understandable even if it wasn't a commercial success. Clover Trail, in contrast, doesn't offer anything new, or even anything that Intel is likely to re-use in future chips.
Sometimes, Microsoft can force Intel to do things. Microsoft Windows runs on most of Intel's processors, and thus Intel must accommodate Microsoft. It's also said that Microsoft holds a blocking patent on the Pentium architecture. In this case, Microsoft needs a "Hail Mary" pass - something that will make Windows 8 desirable on pocket and tablet platforms against IOS and Android, while the desktop market Microsoft built their company upon diminishes. But Clover Trail wasn't a good bet. Ultimately, it will probably sour the Microsoft-Intel relationship.
Linux developers might take it as a
challenge to provide support for Clover Trail. Reverse engineering probably won't be much of a problem.
The next step for the Atom architecture is
Valley View, which does away with PowerVR, replacing it with Intel's own grapics. Intel promises good Linux support for this architecture, and it will use smaller transistors, thus providing a speed and power efficiency gain.
AMD's "Hondo" processor is being marketed as "Windows Only" too. Microsoft must be paying both manufacturers a lot for this. However, AMD is not saying the chip won't run Linux, and is indeed preparing Linux support, though
not for Android. Hondo might make more sense on netbooks, ultra-low-power Linux servers providing SOHO file storage and services, set-top boxes, and empowered TVs.
Update: Intel announced on Monday 17-September-2011 that they'd port Android and Linux to some version of Clover Trail - but didn't state any specifics. So, they have backed down somewhat from the "no Linux" statement. This seems to be a reaction to the annoyance of the Linux community, which does sell lots of Intel servers. Perhaps this all started as a single Intel marketing guy getting off-message? But the chip will still have the PowerVR graphics, which remains less-than-optimal. I'll wait for Valley View.
