Re: POWER vs Intel for DB loads

From: Karl Arao <karlarao_at_gmail.com>
Date: Sat, 10 Aug 2013 00:49:54 +0800
Message-ID: <CACNsJncbK=ihvzr4QwKzpATAqx_zT+JvxsF29KWZxuzkjaviFA_at_mail.gmail.com>

oh my bad, correction, I was talking about the SPECint_rate2006 and that's also on the title of this link http://goo.gl/xj7o8 if these are the machines you are getting, then yes Power is the winner

• P750 Karl-MacBook:SPECint_rate2006 karl$ cat spec.txt | grep -i 750 | sort -nk1 | grep -i ibm | grep -v Intel | grep -v SLES | grep -v Red 50, 32, 8, 4, 4, 1150, 1600, IBM Corporation, IBM Power 750 Express (3.5 GHz 32 core), Feb-13 54.375, 32, 8, 4, 4, 1230, 1740, IBM Corporation, IBM Power 750 Express (4.0 GHz 32 core), Feb-13
• E5 Karl-MacBook:SPECint_rate2006 karl$ cat spec.txt | grep -i intel | grep -i "E5-26" | sort -nk1 | grep -i sun | grep -v Blade 29.8125, 16, 2, 8, 2, 455, 477, Oracle Corporation, Sun Netra X4270 M3 (Intel Xeon E5-2658 2.1GHz), Apr-12 43.875, 16, 2, 8, 2, 628, 702, Oracle Corporation, Sun Fire X4170 M3 (Intel Xeon E5-2690 2.9GHz), Apr-12 43.875, 16, 2, 8, 2, 628, 702, Oracle Corporation, Sun Server X3-2 (Intel Xeon E5-2690 2.9GHz), Jul-12 44.0625, 16, 2, 8, 2, 630, 705, Oracle Corporation, Sun Fire X4270 M3 (Intel Xeon E5-2690 2.9GHz), Apr-12 44.0625, 16, 2, 8, 2, 630, 705, Oracle Corporation, Sun Server X3-2L (Intel Xeon E5-2690 2.9GHz), Jul-12

On Sat, Aug 10, 2013 at 12:20 AM, Rich Jesse < rjoralist2_at_society.servebeer.com> wrote:

> Karl writes:
>
> > While you don't have the machine yet.. you can do all these comparisons
> > using specint_rate2006, depends on what P750 you are getting then compare
> > it to your intel hardware of choice. Here's the data points
>
> Thanks Karl and others for the spec suggestion. I did look at that, but
> missed the POWER 750 because IBM didn't supply numbers for specint2006,
> just
> specint_rate2006. Curious...unless specint2006 is just easy to calculate
> (rate/cores)?
>
> I also looked at TPC even though I feel it's heavily skewed by unreasonable
> configurations. But no POWER 750 for C, E, or H.
>
> That's OK. The specint_rate2006 seems to lean (heavily!) toward POWER.
>
> Thanks again!
>
> Rich
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

-- 
Karl Arao
Blog: karlarao.wordpress.com
Wiki: karlarao.tiddlyspot.com
Twitter: _at_karlarao <http://twitter.com/karlarao>


--
http://www.freelists.org/webpage/oracle-l

Received on Fri Aug 09 2013 - 18:49:54 CEST