rbw at ahpcrc.org
Wed Mar 12 08:48:58 PST 2003
On, Wed, 12 Mar 2003 10:19, Robert Brown:
>> > 200/5600 = .0357 or 3.57% (so 5% is a gift)
>> oh, I see. to me, that's a strange definition of "peak",
>> since stream is, by intention, always bottlenecked on
>> memory bandwidth, since its FSB is either 3.2 or 4.3 GB/s.
>> it'll deliver roughly 50% of that to stream.
>Stream also generally does not deliver worst case memory bandwidth --
>far from it. A "streaming" pass through memory allows all sorts of
>prefetch optimizations to be done that partially parallelize the memory
>access times, as does e.g. SDRAM's basic architecture (which
>accomplishes similar parallelization on a smaller scale).
True. Stream gives the most vector-like view of non-vector ISA's like
IA-32 and such. Pre-fetch is "just another word for nothing we can
vector load." I am tossing dwarf vectors (SSE, SSE2, etc.). No pun-y
And, I give a catechistic node of approval to Robert's ...
More information about the Beowulf