anon (anon.delete@this.anon.com) on August 11, 2014 4:18 am wrote:
> Brett (ggtgp.delete@this.yahoo.com) on August 10, 2014 11:45 pm wrote:
> > anon (anon.delete@this.anon.com) on August 10, 2014 9:03 pm wrote:
> > > Brett (ggtgp.delete@this.yahoo.com) on August 10, 2014 2:21 pm wrote:
> > > > Michael S (already5chosen.delete@this.yahoo.com) on August 10, 2014 2:21 am wrote:
> > > > > Now you are out of your depth. Even dead Itanium is going to last for at least 5-6
> > > > > years. Such live bodies as SPARC and esp. Power will be around for much longer.
> > > >
> > > > Zombies do not count, Freescale is still selling Motorola 68000 chips for Gods sake.
> > >
> > > How would the POWER business model, that you say is unsustainable, become
> > > sustainable if it was switched to a new, unproven, and incompatible ISA?
> > >
> > > At this point, if IBM is to persist with custom high end server CPU
> > > design, they seem far better in continuing with the OpenPower push.
> >
> > The ARM bandwagon is sucking all the oxygen from OpenPower, everyone is converting designs to ARM.
>
> Who is? Nobody has anything equivalent to POWER8.

I agree, that is my point as well.

> You should count the number of enterprise operating systems available to POWER and for
> ARM. Then count the number of databases. Then the number of business middleware. etc.

AMD is selling a fair number of ARM64 Seattle servers to developers to port software.
These software houses would LOVE to have some POWER8 ARM64 boxes as well.
The Seattle boxes would get reduced to test servers while programmers did all the work on the faster IBM boxes.

With high and low covered by AMD and IBM the software floodgates would open creating a full server market in short order. The five year plan instead of the ten year attack only from the bottom approach.

Lets look at this from IBM's point of view, projecting a mere 10 years into the future.
Stay on PowerPC, it is a hassle to move from x86 or ARM64 to POWER, the market is split three ways, time to shut down the business.
Dominate the high end of the ARM64 market, have people coming to you wanting your faster chips, to run their existing binaries on.

Does IBM think either of these two choices make sense long term, or just shut the division down slowly.