<strong>I remember programming in FORTAN on a CRAY. It was one of the best vectorizing compilers available. Even then, you could get up to a 2x improvement by coding it using the vector primitives yourself instead of letting the compiler do it.
Right... but I don't think enough research has been done to say that its not possible. We take a lot of things for granted today that were deemed "not possible" 20 years ago.
[quote] Also there was a rumor linked with this Sahara chip a while back that said the whole chip would somehow produce Altivec-like acceleration as opposed to co-processor. Any1 know anything about this? <hr></blockquote>
I also remember an IBM spokesman saying this some ~7 months ago.
If IBM can produce even 'near-SIMD performance' without special coding -- WOW!
Right... but I don't think enough research has been done to say that its not possible. We take a lot of things for granted today that were deemed "not possible" 20 years ago.</strong><hr></blockquote>
I hope that you are right and that we will see a language that is the anwer to the problem in the same way LINDA is an answer to distributed computing.
There seems to be ongoing work to make gcc vectorize altivec code (-maltivec)
[quote] Aldy Hernandez, of Red Hat, Inc has contributed extensions to the PowerPC port supporting the AltiVec programming model (SIMD). The support, though presently useful, is experimental and is expected to stabilize for 3.2. The support is written to conform to Motorola's AltiVec specs.
if you were apple, and IBM had a monster PPC-compliant CPU, wouldn't you persuade them to get into the blade and workstation market with it?
Get my drift...</strong><hr></blockquote>
mslee, are suggesting IBM will push th Power series down-market or that IBM may become a reseller of Apple machines as their bottom end (with IBM chips of course)...? <img src="graemlins/surprised.gif" border="0" alt="[Surprised]" />
Comments
<strong>I remember programming in FORTAN on a CRAY. It was one of the best vectorizing compilers available. Even then, you could get up to a 2x improvement by coding it using the vector primitives yourself instead of letting the compiler do it.
C is just not easy to vectorize. There are attempt to solve this with new languages (i.e. <a href="http://dynamo.ecn.purdue.edu/~hankd/SWAR/Scc.html)," target="_blank">http://dynamo.ecn.purdue.edu/~hankd/SWAR/Scc.html),</a> but performance is so tied to the actual hardware, that it may not be possible.</strong><hr></blockquote>
Right... but I don't think enough research has been done to say that its not possible. We take a lot of things for granted today that were deemed "not possible" 20 years ago.
[quote] Also there was a rumor linked with this Sahara chip a while back that said the whole chip would somehow produce Altivec-like acceleration as opposed to co-processor. Any1 know anything about this? <hr></blockquote>
I also remember an IBM spokesman saying this some ~7 months ago.
If IBM can produce even 'near-SIMD performance' without special coding -- WOW!
Oh, and double-precision please.
<strong>
Right... but I don't think enough research has been done to say that its not possible. We take a lot of things for granted today that were deemed "not possible" 20 years ago.</strong><hr></blockquote>
I hope that you are right and that we will see a language that is the anwer to the problem in the same way LINDA is an answer to distributed computing.
There seems to be ongoing work to make gcc vectorize altivec code (-maltivec)
[quote] Aldy Hernandez, of Red Hat, Inc has contributed extensions to the PowerPC port supporting the AltiVec programming model (SIMD). The support, though presently useful, is experimental and is expected to stabilize for 3.2. The support is written to conform to Motorola's AltiVec specs.
<hr></blockquote>
Also interesting while we are at it (from <a href="http://gcc.gnu.org/gcc-3.1/changes.html"; target="_blank">gcc 3.1 additions</a>)
[quote]The PowerPC back end has added 64-bit PowerPC GNU/Linux support<hr></blockquote>
[ 06-23-2002: Message edited by: *l++ ]</p>
<strong>the IBM rumor is legit, i think.
if you were apple, and IBM had a monster PPC-compliant CPU, wouldn't you persuade them to get into the blade and workstation market with it?
Get my drift...</strong><hr></blockquote>
mslee, are suggesting IBM will push th Power series down-market or that IBM may become a reseller of Apple machines as their bottom end (with IBM chips of course)...? <img src="graemlins/surprised.gif" border="0" alt="[Surprised]" />