By: Aaron Spink (aaronspink.delete@this.notearthlink.net), August 10, 2014 9:19 am
Room: Moderated Discussions
Michael S (already5chosen.delete@this.yahoo.com) on August 10, 2014 8:48 am wrote:
> If I am not mistaken, when compared to its main server competitors, i.e. Power and SPARC, gcc-all-around
> comparisons are more favorable to Intel (x86, not Itanium) than vendor's compiler vs vendor's compiler.
>
> Also, among those 3 vendors + AMD, there appears to be good correlation between official SPECInt_rate
> scores and SAP SD 2-tier scores. Back when TCP-C was not dead, there was not a bad correlation between
> it and SPECInt_rate as well, except for big-core-count Itanium vs big-core-count IBM Power.
>
This is all basically true. Its important to point out that anything running large DBs, etc, is generally heavily optimized and in line with vendor compilers for comparison purposes. Its also worth pointing out that Spec2006 will overall give advantages to lower end processors than high end ones since the dataset sizes are out of date. Most of the higher end processors are targeting significantly larger dataset sizes than are used in Spec2006.
> If I am not mistaken, when compared to its main server competitors, i.e. Power and SPARC, gcc-all-around
> comparisons are more favorable to Intel (x86, not Itanium) than vendor's compiler vs vendor's compiler.
>
> Also, among those 3 vendors + AMD, there appears to be good correlation between official SPECInt_rate
> scores and SAP SD 2-tier scores. Back when TCP-C was not dead, there was not a bad correlation between
> it and SPECInt_rate as well, except for big-core-count Itanium vs big-core-count IBM Power.
>
This is all basically true. Its important to point out that anything running large DBs, etc, is generally heavily optimized and in line with vendor compilers for comparison purposes. Its also worth pointing out that Spec2006 will overall give advantages to lower end processors than high end ones since the dataset sizes are out of date. Most of the higher end processors are targeting significantly larger dataset sizes than are used in Spec2006.