Re: Intel enters in smartphone market: Game Over for ARM.
 

game on, more like it.

Re: Intel enters in smartphone market: Game Over for ARM.
 

Huh? Isn't that what I just wrote? Why do you have to quote me?

Re: Intel enters in smartphone market: Game Over for ARM.
 

For me, the important thing about this is that Intel sells chipsets, not devices.

With the atom chipset we've seen tablets, netbooks and other devices from more vendors you can think of, thanks to chinese manufacturers.

You can mount this Medfield thing on any thing you can think of, and you can put around it what you want, regarding companion chips for communications, networking or whatever.

You can start using it to sell educative computers for schools, like, for example, the Neo.

Re: Intel enters in smartphone market: Game Over for ARM.
 

A keyboard with tiny black and white text only display for 169$ ? And it doesn't seem to actually do anything. I'm underwhelmed! :D

For 25$ you can get the incomparably more powerful Raspberry Pi board and actually do something useful with it. Obtaining a second hand monitor and keyboard is left as an exercise for the reader, but should not be an issue.

Re: Intel enters in smartphone market: Game Over for ARM.
 

That is sadly not true, because Medfield does not necessarily come with BIOS or EFI. Or even a PCI bus. Windows would hardly run and GNU/Linux x86 would need a new "subarchitecture".

Re: Intel enters in smartphone market: Game Over for ARM.
 

Still, storing 40 "second hand monitors" in the class cupboard, or moving from class to class carying a 4 kg monitor (with a power cord..) just to take notes or write answers to a test does not look like an overwhelming idea to me either ;)

Re: Intel enters in smartphone market: Game Over for ARM.
 

But if the board runs say Android, does that ensure that the necessary components for a standard Linux are there?

Re: Intel enters in smartphone market: Game Over for ARM.
 

Rather than the components for a standard Android are there.... exactly as what is happening now on ARM, you know.

Re: Intel enters in smartphone market: Game Over for ARM.
 

For me a Intel smartphone would be cool because it would make running your own version of a distribution a little bit easier.
Desktop distribution don't always support ARM yet, and when they do it usually isn't that well tested on the diverse hardware available. Also compiling packages in the native architecture is easier and faster, and having a Core i7 Desktop PC to compile anything for my phone does make that a lot more comfortable.

That said, running desktop software on a smartphone isn't always a good idea. For a proper user experience interactive applications mostly need to be rewritten with a different GUI. Also non-interactive software needs to be optimized for power consumption. That is something that works a bit differently in ARM and atom as far as I know (ARM puts entire parts of the CPU in idle state, Atom only throttles down afaik).

The most interesting aspect is the performance. The atom CPU performs very well in single threaded applications, but the current dual core Cortex A9 chips beat intel when multiple cores can be used or applications that use the GPU or can use a DSP (video decoding).

I know Intel will have multicore Atom on the way and improved GPU, but so does ARM with Cortex A15 and PowerVR series 6.
Let us see what the actual hardware will cost first :)

Re: Intel enters in smartphone market: Game Over for ARM.
 

Most obvious mistake in the Extremetech article is that he is assuming that competition stands still while Intel continues to develop better and better chips.

He also did not mention that some Android applications won't even work on x86 without recompiling. Most will probably work but applications that are using native code - such as some graphics intensive games - simply won't work. This will create even more fragmentation as already on my Galaxy S2 I cannot see all games on the market, for example (many require a Tegra2).

He also claims that chip manufacturers just license the core from ARM and churn out identical chips based on them. This is not true, in fact there can be big performance differences between two application processors even if they share the same core. For example, in OMAP4 there are two 32-bit DDR interfaces vs one on Tegra2.

He also claims that there is no competition today and that x86 will bring that into mobile space. This is also not true. Are TI, Qualcomm, ST-Ericsson, Nvidia, Samsung & the rest all part of the same happy family? I don't think so! I have been working in the electronics industry for quite some time and the price of ARM-based application processors has been steadily going down. Already today you can get a Cortex-A8 based processor at below 10USD even in small volumes.

I don't dare predict what kind of market share Intel will be able to command in mobile space but I'm quite sure that it will not obliterate ARM in a matter of few years as the linked article suggests.