[Beowulf] NVIDIA buying PGI

[Joe Landman]

On 08/01/2013 05:02 PM, mathog wrote:
> OJoe Landman wrote:
>
>> It makes sense from a competitive "gird your loins" sort of way, as
>> Intel has its compiler happily able to emit Larrabee ... er ... MIC
>> executables.  Is that technology anticompetitive in Intel's hands?
>
> Yes.  When hardware companies control compilers the end user suffers
> in that, inevitably, competition is reduced when the hardware company
> causes one or more
> of the compiler's target  hardware platforms to either be desupported,
> as here, or under supported, or
> never developed, in the case of new hardware.

So ... you are making the presumption that this reduction in compiler 
target will occur.

Your argument amounts to "once something is created and put out in the 
market, it is anticompetitive to change it in such a way as to reduce 
support for one possible platform, irrespective of increases of support 
in others."

Am I reading this correctly?  Are you making the claim that features and 
targets must be carried *forever* (as in until the heat death of the 
universe) lest they be deemed "anti-competitive"?

Ok, quickie poll on Beowulf:

a) whom is running Intel and/or AMD kit?
b) whom is considering running new AMD kit?
c) whom would reconsider if PG end-of-lifed (which I think is Dave's 
speculation) the AMD targeted compiler?
d) whom doesn't really care about PG's eol status, and it wouldn't 
impact purchase decisions in either direction?

Is not the answer to "c" precisely the behavior you are afraid of?

I suspect that the answer to "d" would be substantially numerically 
larger than "c".  Which would somewhat ... er ... diminish ... the claim 
of anti-competitive behavior influencing the market.

I thought your pilot analogy was funny for a number of reasons, not the 
least of which is that unions seek to become monopoly suppliers of their 
wares, and control the negotiation process to their sole advantage. 
That is, they seek to usurp the normal competitive market pressures. 
Hardly a fit analogy for this.

If my read on your argument is correct, you are generally non-supportive 
of reducing platform support in general, and specifically for reasons of 
acquisition which you deem to be anti-competitive.  Anti-competitive 
means that a customer or market place would be injured due to actions of 
the entity making those decisions and performing those actions.  What 
specific injury can you claim when there are *an abundance* of vendors 
and open alternatives in this space?

I guess I am somewhat incredulous over this, given that OpenMP 4.x was 
just released, which includes much of what PG has already pioneered, so 
I expect these capabilities to show up in gcc, LLVM, Intel, Open64, 
PathScale, NAG, .... ad nausem.  And I guess I am just missing the point 
completely ... because I see lots of choice, lots of options, and no 
significant loss to the market space.   No massive reduction of choice. 
  No sudden requirements for massive additional costs for users.  No 
monopoly forming (which would be anticompetitive).


-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics, Inc.
email: landman at scalableinformatics.com
web  : http://scalableinformatics.com
        http://scalableinformatics.com/siflash
phone: +1 734 786 8423 x121
fax  : +1 866 888 3112
cell : +1 734 612 4615