Larger L2, but no increase in latency?
When Prescott first got a 2MB L2 cache, we noticed that along with a larger L2 came a 17% increase in access latency.  The end result was a mixed bag of performance, with some applications benefitting from the larger cache while others were hampered by the increase in L2 latency.  Overall, the end result was that the two performance elements balanced each other out and Prescott 2M generally offered no real performance improvement over the 1MB version. 

With Presler, each core also gets an upgraded 2MB cache, as compared to the 1MB L2 cache found in Smithfield.  The upgrade is similar to what we saw with Prescott, so we assumed that along with a larger L2 cache per core, Presler's L2 cache also received an increase in L2 cache latency over Smithfield. 

In order to confirm, we ran ScienceMark 2.0 and Cachemem:

   Cachemem L2 Latency (128KB block, 64-byte stride)  ScienceMark L2 Latency (64-byte stride)
AMD Athlon 64 X2 4800+ 17 cycles 17 cycles
Intel Smithfield 2.8GHz 27 cycles 27 cycles
Intel Presler 2.8GHz 27 cycles 27 cycles
Intel Prescott 2M 27 cycles 27 cycles
Intel Prescott 1M 23 cycles 23 cycles

What we found was extremely interesting; however, Presler does have the same 27 cycle L2 cache as Prescott 2M, but so does Smithfield.  We simply took for granted that Smithfield was nothing more than two Prescott 1M cores put together, but this data shows us that Smithfield actually had the same higher latency L2 cache as Prescott 2M.  

Although we were expecting Presler to give us a higher latency L2 over Smithfield, it looks like Smithfield actually had a higher latency L2 to begin with.  This means that, at the same clock speed, Presler will be at least as fast as Smithfield, if not faster.  Normally, we take for granted that a new core means better performance, but Intel has let us down in the past; luckily, this time we're not put in such a situation. 

Literally Dual Core Presler vs. Smithfield - A Brief Look
Comments Locked

84 Comments

View All Comments

  • Betwon - Friday, December 30, 2005 - link

    edit:
    Why is it slower the latency of the memory? 101ns is much more than 5x ns. where is the 'on-die' communication? Your test program is wrong?
  • Viditor - Friday, December 30, 2005 - link

    Thanks Anand!

    I don't know if you'll have time, but one of the things lacking in all of the other reviews of the OC XE955 is a comparison to an OC X2 4800...
    Speculation is quite rife, and it would be a good comparison IMNSHO.

    Cheers!
  • Gigahertz19 - Friday, December 30, 2005 - link

    Intel's back...back again...backkkkkkk again..backkkkkkkk again...du dah duh da
  • yacoub - Tuesday, January 3, 2006 - link

    If by "back" you mean finally (barely) able to compete with existing AMD performance, then yes. ;P

Log in

Don't have an account? Sign up now