David Kanter (dkanter@realworldtech.com) on 11/9/07 wrote:
---------------------------
>I just finished up my preview of Harpertown, and nary a moment too soon. It's only
>a preview if it gets up before the product is released, so I managed it at the last moment.
>
>This was the hardest review I've done yet...largely because I had to evacuate my
>testing facility about half way through the testing.
>
>Anyway, take a look - there is actually some relatively interesting analysis that
>I have to use in order to estimate the IPC improvement from Clovertown to Harpertown.
>I had two processors of different frequencies, and all things being equal a 2.33GHz
>processor will have a higher IPC than a 3GHz one...which made the estimates rather tricky.
>
>http://www.realworldtech.com/page.cfm?ArticleID=RWT110907063310
>
>Take a read and feedback is always welcome.
>
>Regards,
>
>David

Thanks for writing up the article.

It's not clear to me how much of the performance differential is due to Harpertown vs. Clovertown, and how much is Stoakley vs. Bensley. Do you have any insights?

A pure Stoakley vs. Bensley review would be interesting. Maybe you could put the Harpertown chip in the Bensley platform and see how it performs?