[FoRK] Fwd: Microsoft Posts First Quarterly Loss Ever [lost decades]

Stephen D. Williams sdw at lig.net
Thu Jul 19 19:32:46 PDT 2012

X86 won't necessarily go away.  Outside of NVidia, it's the fastest thing out there if you don't care about power, heat, or cost.  
While it is a complicated, messy mess, it works well now and has many tricks that are not found elsewhere.

On the other hand, ARM and associated GPU cores are evolving quickly and have vastly lower power usage.  Atom is interesting, and 
Intel has invested in some other technology, but the ARM related vendors have a pretty good head start.  If they figure out the 
processor tricks that make x86 so fast, and it can be done in a low power mass market mobile chip, Intel is in a little trouble.  
Still a horse race and one that no one cares too much about.  Android already supports many different CPU architectures at the same 
time in shipping executables, so it's a non-issue of any significance.  Already you can make an Android app that uses several ARM 
versions and x86 and ship it now.  Android x86 was in limbo for a while, but now is, I understand, fully supported with ICS on 
Google TV devices.


On 7/19/12 7:09 PM, Gregory Alan Bolcer wrote:
> Okay, I agree with everything you are saying, but
> How long until you think x86 goes away?
> Greg
> On 7/19/2012 6:44 PM, Stephen Williams wrote:
>> MicroWho?
>> ...
>> The only question is how quickly Windows and Office will fade or be
>> upstaged.  I think I know one interesting path, but we'll see.
>> Microsoft could easily pivot and recover, but they've shown every sign
>> of being too thick-headed and thick-bellied to do anything serious about
>> it.  And we're probably better off for it.  I could solve their
>> problems, but they wouldn't listen even if they heard me, any more than
>> McNealy listened by my suggestions a decade ago.
>> All that said, this is just a blip from a write-down apparently. MS has
>> plenty of cash, still has a lot of suckers hooked, and has plenty of
>> maneuvering room.  They could surprise, but past performance leads to
>> the conclusion that it wouldn't be a nice surprise.
>> sdw

More information about the FoRK mailing list