Interesting, old-style Opterons doing extremely well in Spec 2006

This is prior to the introduction of Barcelona architecture Opterons,
but it looks like even these old codgers are doing extremely well in the
SpecInt_rate2006 and SpecFP_rate2006 benches. I wonder why?

TG Daily - Opteron vs. Xeon performance battle goes into the next round
"In fact, the numbers do look good on AMD’s presentation. The 3.0 GHz
Opteron 2222SE outpaces Intel’s Xeon 5160 in three server-specific
benchmarks (SPECint_rate_2006, SPECint_rate2006, SPECompM2001) by up to
24%. In other benchmarks, such as Fluent, the advantage approaches 40%.
"
http://www.tgdaily.com/content/view/31489/135/

--
There is no failure, only delayed success

Yousuf Khan wrote:

> This is prior to the introduction of Barcelona architecture Opterons,
> but it looks like even these old codgers are doing extremely well in the
> SpecInt_rate2006 and SpecFP_rate2006 benches. I wonder why?
>
> TG Daily - Opteron vs. Xeon performance battle goes into the next round
> "In fact, the numbers do look good on AMD’s presentation. The 3.0 GHz
> Opteron 2222SE outpaces Intel’s Xeon 5160 in three server-specific
> benchmarks (SPECint_rate_2006, SPECint_rate2006, SPECompM2001) by up to
> 24%. In other benchmarks, such as Fluent, the advantage approaches 40%.
> "
> http://www.tgdaily.com/content/view/31489/135/
>

interesting.

makes me feel good about my Opty Troy 250s.

i see that newegg still offers roughly 21 of the
socket 940 CPUs. too bad i didn't wait till now
to get them.

i mean, i had a pair of 248s, but wanted to
move my older Tyan up to the 248 (turned out
it didn't make that much diff; but moving to
a faster boot drive and a raid0 "second"
drive did)

the thing i learned with the two mobos is that
the chipset is just as important as the CPU
(they (N/S/bridge) run *hot* on my more recent
SuperMicro, but it is the noticeably faster machine;
easy enough to just drop a loose 80mm fan into the
bottom of the case blowing on the N/S/bridge)

on a side note: any idea if George (McD) and Keith
are still alive?

bill

willbill <trek@worldwide.net> wrote:
: Yousuf Khan wrote:
:
<snip>

: on a side note: any idea if George (McD) and Keith
: are still alive?

Have been wondering the same thing. The number of MIA regulars in this NG
is just astounding!

/TLG

On Apr 6, 12:45 pm, willbill <t...@worldwide.net> wrote:

>
> on a side note: any idea if George (McD) and Keith
> are still alive?
>

I've seen posts from Keith on comp.arch.

Robert.

Robert Myers wrote:
> On Apr 6, 12:45 pm, willbill <t...@worldwide.net> wrote:
>
>
>>on a side note: any idea if George (McD) and Keith
>>are still alive?
>>
>
>
> I've seen posts from Keith on comp.arch.
>
> Robert.
>

thanks robert. i know i could have tried
google to check on any recent posting but
i seldom do that and never think of it.
i don't know either of them, but i miss
their savvy posting

bill

On Thu, 05 Apr 2007 23:58:26 -0400, Yousuf Khan <bbbl67@yahoo.com>
wrote:

>This is prior to the introduction of Barcelona architecture Opterons,
>but it looks like even these old codgers are doing extremely well in the
>SpecInt_rate2006 and SpecFP_rate2006 benches. I wonder why?
>
>TG Daily - Opteron vs. Xeon performance battle goes into the next round
>"In fact, the numbers do look good on AMD’s presentation. The 3.0 GHz
>Opteron 2222SE outpaces Intel’s Xeon 5160 in three server-specific
>benchmarks (SPECint_rate_2006, SPECint_rate2006, SPECompM2001) by up to
>24%. In other benchmarks, such as Fluent, the advantage approaches 40%.
> "
>http://www.tgdaily.com/content/view/31489/135/

Intel's kind of got a point that the results are 'rubbish', given that
they're comparing AMD's latest and greatest dual-core chip to a
processor Intel released almost a year ago (May 2006). Intel's newest
quad core chips easily outpace the AMD chips in most of those
benchmarks.

Still, it does show some promise for AMD's quad-core "Barcelona"
Opterons. Even if they only keep the status quo then they should
often be competitive with Intel's server offerings. If the AMD chips
deliver the promised noticeable boost in per-core performance
(relative to current Opterons) then AMD could well be back in the
drivers seat.
--
Tony Hill
hilla <underscore> 20 <at> yahoo <dot> ca