Tom Marchant wrote:

<begin extract>
In this way, at least some of the time that the processor is idle
waiting for data from memory, it is able to do useful work.
</end extract>

and I should reformulate this as

In this way at least some of the time that the processor would
otherwise be idle waiting for data from memory needed by the current
thread it is able to do useful work on another thread.

There is reason to believe that this is|could be the case for more
than 40% of this memory-wait time; implemented well, SMT may thus
improve performance significantly.

Tom's original post makes this point clearly and helpfully.  This one
is only a minor clarification.


John Gilmore, Ashland, MA 01721 - USA

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to