By: someone (someone.delete@this.somewhere.com), November 19, 2010 7:52 am
Room: Moderated Discussions
Doug Siebert (foo@foo.bar) on 11/19/10 wrote:
---------------------------
>someone (someone@somewhere.com) on 11/18/10 wrote:
>---------------------------
>>I guess you are ignorant of the fact that IA64 includes
>>means for the compiler to associate hints about the
>>spacial and temporal locality of the data associated
>>with each load and store instruction as well as specify
>>the degree of ordering strictness required between sets
>>of memory transfers. These features, AFAIK unique to
>>IA64, allow conflict and capacity misses to be reduced
>>as well as avoiding stalls from hardware enforcement
>>of unnecessary memory ordering.
>>
>>It is funny how negative opinion of IA64 is so highly
>>correlated with degree of sheer ignorance about it. ;^)
>
>
>Are you suggesting that one's opinion of IA64 should be governed by its unique
>features, and since it does a lot of things differently from anyone else we should have a positive opinion about it?
I am simply commenting that IMO the people who hold the
strongest negative opinions about IPF in forums are often
the most ignorant about details of the architecture itself.
The previous poster's comments are an excellent example.
>
>Funny, I tend to form my opinions about IA64 versus x86 versus POWER versus SPARC
>based on real world factors such as their relative performance, power usage, cost,
>etc. (taking into account x86's obvious advantages in economies of scale where cost
>is concerned) and my assessment of the prospects for improvements to these factors in the future.
As it should be for the big picture view. I too am very
critical about some aspects of IA64 but only where it
is deserved - a few early design mistakes and repeated
instances of very poor product delivery execution that
have significantly hurt, to a large extent permanently,
the scope and extent of IPF's commercial success.
However in a technical discussion of the relative merits
of EPIC/IA64 vs traditional architectures one should at
least be generally informed of what IA64 is and isn't
all about. IMO most negative opinions about IPF at a
technical level is based on myths/misunderstandings
endlessly recycled, often by those with an ideological
predisposition to dislike Intel and its products.
---------------------------
>someone (someone@somewhere.com) on 11/18/10 wrote:
>---------------------------
>>I guess you are ignorant of the fact that IA64 includes
>>means for the compiler to associate hints about the
>>spacial and temporal locality of the data associated
>>with each load and store instruction as well as specify
>>the degree of ordering strictness required between sets
>>of memory transfers. These features, AFAIK unique to
>>IA64, allow conflict and capacity misses to be reduced
>>as well as avoiding stalls from hardware enforcement
>>of unnecessary memory ordering.
>>
>>It is funny how negative opinion of IA64 is so highly
>>correlated with degree of sheer ignorance about it. ;^)
>
>
>Are you suggesting that one's opinion of IA64 should be governed by its unique
>features, and since it does a lot of things differently from anyone else we should have a positive opinion about it?
I am simply commenting that IMO the people who hold the
strongest negative opinions about IPF in forums are often
the most ignorant about details of the architecture itself.
The previous poster's comments are an excellent example.
>
>Funny, I tend to form my opinions about IA64 versus x86 versus POWER versus SPARC
>based on real world factors such as their relative performance, power usage, cost,
>etc. (taking into account x86's obvious advantages in economies of scale where cost
>is concerned) and my assessment of the prospects for improvements to these factors in the future.
As it should be for the big picture view. I too am very
critical about some aspects of IA64 but only where it
is deserved - a few early design mistakes and repeated
instances of very poor product delivery execution that
have significantly hurt, to a large extent permanently,
the scope and extent of IPF's commercial success.
However in a technical discussion of the relative merits
of EPIC/IA64 vs traditional architectures one should at
least be generally informed of what IA64 is and isn't
all about. IMO most negative opinions about IPF at a
technical level is based on myths/misunderstandings
endlessly recycled, often by those with an ideological
predisposition to dislike Intel and its products.