gallier2 (gallier2@gmx.de) on 12/20/11 wrote:
---------------------------
[snip]
>How many $1K chips does Intel sell in comparison to <$200
>chips? Yeah, not much. Why would it be better for AMD to
>try to sell to that dwindling demographic (gamers
>with more cash than brain)?

Not that the numbers would work out this way nor that
Intel would be likely to allow AMD such a foothold, but
if AMD had 15% of the market by revenue and effectively
none of the high end, then gaining 90% of the top 1% of the
market with 10 times greater average selling price of
AMD's current overall ASP would allow a AMD to lose 40% of
it revenue in other chips (whether from lower ASPs or
lower volume).

In addition, there might be some boost in the brand value
among the extreme gamer wannabe population (who might buy
a less overpriced CPU and recommend AMD to less technical
friends--not based on actual value but 'coolness') as well
as some of the market in general. AMD did afterall go to
the effort to get a Guinness World's Record for CPU clock
frequency.

>Don't forget, the good enough is the worst brake in CPU
>developement. To give an example, I have an AMD Phenom II
>720 in my PC and I don't need a (CPU) upgrade anytime
>soon. More disk, maybe SSD, a second screen, but CPU
>power, certainly not. For the few CPU intensive tasks I
>happen to do (video transcoding), it makes no difference
>if it takes 30min or 4 hours.

With the emphasis on batter-powered devices, it is not
clear that Microsoft and its associated ISVs can develop
software that is perceived to be necessary and can
adequately drain the CPU performance glut, especially with
GPU offload of DSP-like tasks (possibly speech recognition).

If one could reuse the display of laptops as one can a
desktop, I suspect that lower-end systems would be
attractive to a larger fraction of consumers. (Though
looking at how uncommon small form factor desktops are,
this kind of change may be unattractive to the mainstream
OEMs. The increase in sales volume may not compensate for
the significant drop in ASPs.) Even for laptops that are
typically plugged in, battery life is not likely to exceed
four years (I am guessing), so that relatively expensive
component could probably not be reused.

(Side rant: I am disappointed that inexpensive,
low-capacity, moderate performance solid state storage is
not broadly available. The emphasis on capacity and
performance--while perhaps appropriate initially--seems to
be ignoring the scalability aspect of sold-state [having a
relatively low fixed cost component compared to disk] and
the power and form-factor [and to a lesser extent now,
noise] advantages.)

Of course, people (including me) are hesitant to purchase
re-manufactured systems (if one has to buy an extended
warranty to feel comfortable, one is losing much of the
price savings and there is the inconvenience of sending in
the computer for repair/replacement, especially if that
means being without a computer for weeks) and a quick cheap
system upgrade service does not seem to be provided
(partially because systems tend not to be designed for
repair--form factor and manufacturability are more
important--and partially because this would be a new
service that might eat out profits from established
sellers).

A rent-to-own business might be able to enter this niche.
Such a business could provide a temporary computer while a
system is in for repairs (or even if a system upgrade will
take a significant amount of time)--having a local physical
presence can be helpful.