List:General Discussion« Previous MessageNext Message »
From:Andy Smith Date:April 23 2009 7:25am
Subject:Re: Mysql on Ultrasparc T2 and floating point performance
View as plain text  
>>
>> Hi Dan,
>>
>> Yes, actually, I already know that parallel performance will be much
>> improved, however, I was expecting more improvement on single threads as
>> well, since the specs say that it takes 40 clock cycles just to access the
>> FPU on the T1, but something like 6 clock cycles on the T2.  So just from
>> that perspective it seems like there should be a significant improvement for
>> single threads, not just parallel performance.  At least that's the way I
>> read the docs from Sun on this.  At any rate, my expectations here are
>> clearly wrong, and I guess I'd just like a better understanding of why I'm
>> getting it wrong.


Hi Rod,

   I wouldn't expect any noticable difference running a single query  
that is taking anthing than less than a second if the only technical  
difference is that it takes 34 less clock cylces to access the FPU,  
thats 34 clock cycles on a cpu with <1GHz clock. The additional FPUs  
and reduced clock cycles to access the FPUs simply improve preformance  
when running mutliple threads that require FPU calculations, ie almost  
no overhead up to 8 simultaneous FPU calculations and less overhead  
when the number of simultaneous calculations goes over the number of  
physical FPUs due to the reduced clock cycles to access a FPU.
In general my understanding of the T1 vs T2 architectures is that the  
CPU and FPU cores are essentially the same regarding preformance per  
thread but with additional threads per core and additional FPUs and  
some other nice bits like on board Gig Ethernet and support for 2  
socket servers, plus any bump in the clock speed over the older boxes  
obviously.

thanks Andy.
Thread
Mysql on Ultrasparc T2 and floating point performanceRod Heyd22 Apr
  • Re: Mysql on Ultrasparc T2 and floating point performanceDan Nelson22 Apr
    • Re: Mysql on Ultrasparc T2 and floating point performanceRod Heyd22 Apr
      • Re: Mysql on Ultrasparc T2 and floating point performancemos22 Apr
        • Re: Mysql on Ultrasparc T2 and floating point performanceAndy Smith23 Apr