Sempron 3300+ beats AMD64 3200+ How?

over_c

Distinguished
May 27, 2003
289
0
18,780
I was reading anandtech's review of the new Sempron 3300+. I was surprised to see that in two tests, the Communication Sysmark 2004 <A HREF="http://anandtech.com/cpuchipsets/showdoc.aspx?i=2395&p=2" target="_new">test</A> and the MusicMatch Jukebox 7.10 <A HREF="http://anandtech.com/cpuchipsets/showdoc.aspx?i=2395&p=5" target="_new">test</A> the Sempron beat the AMD64 3200+.

Sempron 3300+______AMD64 3200+
2.0 GHz____________2.0 GHz
128kB L2___________512kB L2
single channel______dual channel memory (same memory used for both tests)
ABIT NF8 nForce3___MSI nForce4 SLI
X800 XT AGP_______X800XT PCIe


How can the Sempron 3300+ beat the AMD64 3200+ at anything? Is it the chipsets, faster cache in the Sempron, or a flawed methodology for anandtech?

I know the AMD64 crushed the Sempron in all other tests. I am just surprised that it lost in any of the tests.
 

Mephistopheles

Distinguished
Feb 10, 2003
2,444
0
19,780
Hm... that is indeed strange...

Well, after all, at least the sempron actually had a 3300 performance rating... so in any case, if it actually did get crushed by an A64 3200, you could complain that the PR system isn't working...
 

over_c

Distinguished
May 27, 2003
289
0
18,780
The performance rating for Semprons is based against Intel Celerons, not P4s. So just as an Intel 3.2 GHz would beat a 3.3 GHz Celeron in most, if not all, applications, so should an AMD64 3200+ beat a Sempron 3300+.
 

TheRod

Distinguished
Aug 2, 2002
2,031
0
19,780
Ok, let's put numbers in perspective...

Communication SYSMark 2004
Sempron 3300+ = 139 (103%)
A64 3200+ = 135 (100%)

MusicMatch Jukebox 7.10
Sempron 3300+ = 531 (102%)
A64 3200+ = 540 (100%)

The Sempron system is based on nForce3 and A64 on nForce4. So they don't use the same platform. And they don't specify if the A64 3200+ is 90nm or not.

<b>From the introduction :</b>
<i>The new Sempron also features all of the enhancements that made it into the 90nm Athlon 64 processors - mainly SSE3 support and some enhancements to the memory controller.</i>

The A64 3200+ is not a Revision E for sure (no SSE3 support nor mem. controller enhancements), this problably explain the Sempron 3300+ good scores in some benchmarks. A64 Rev. E are not far away, they will probably retake the crown in those benchmarks.

-
GA-K8NF-9 / <b><font color=green>Athlon 64 3200+</font color=green> @ 3800+</b>
Infineon DDR400 (CL2.5) 2x512Megs
<font color=green>GeForce 6600GT 128Megs</font color=green>
<A HREF="http://www.getfirefox.com" target="_new">Firefox</A>
 

pat

Expert
Peoples buy number.. not performance.. dual channel is better than single, larger cache is better, Socket 939 has more pins than 754, but often forget to really look at the real thing.

Single channel memory bandwith has more efficiency than dual channel in % of used bandwidth versus available bandwidth. So, less prone to penality hits with multiple acces to memory. Smaller cache is sometime faster than larger one on A64 because of the efficient mem controller and due to the facts that it is less prone to penality hits with flush/refill cache, access is faster because of the smaller size too. Then, it use the improved venice core too. So, no I'm not surprized about the sempron being that fast, but I have rather a deception with the "oh mighty" socket 939 being that slow.

Hey, I miss my 3000+ 754 for its raw speed... @00 MHZ less is hard to compensate with DC.. AMD should have kept the same clock for the 754 and 939, and use a 3100+ 939 vs a 3000+ 754.

<font color=red>Sig space for rent. make your offer.</font color=red>
 

TheRod

Distinguished
Aug 2, 2002
2,031
0
19,780
I think we will always debate about the AMD Performance Rating. It will never be perfect!

But overall, is not that bad... For the average user, the difference in performance between 3000+ and 3200+ is not very significative. At least, AMD try to compare to the competition. On the other hand the new Intel Model System is becoming overcrowed with tons of family and variants...

Is it better to get a 570 or a 630???

-
GA-K8NF-9 / <b><font color=green>Athlon 64 3200+</font color=green> @ 3800+</b>
Infineon DDR400 (CL2.5) 2x512Megs
<font color=green>GeForce 6600GT 128Megs</font color=green>
<A HREF="http://www.getfirefox.com" target="_new">Firefox</A>
 

Patchworkblue

Distinguished
Aug 12, 2003
127
0
18,680
Oh, come on, you guys are smarter than this. You forgot to take into account that the Anandtech reviewers tested the processors by flinging their feces at them and seeing whose crap sticks to what processor longest.
Seriously, I know that THG has been doing a lot of consumer electronics and other fluff lately, but that doesn't mean you should be reading Anandtech. You should be ashamed.

Women--can't live with them, can't have heterosexual same-species intercourse without them.
 

Spitfire_x86

Splendid
Jun 26, 2002
7,248
0
25,780
No, both were clocked at 2.0 GHz, while A64 3200+ had 4x larger L2 cache and twice memory bandwidth

------------
<font color=orange><b><A HREF="http://www.mozilla.org/products/firefox" target="_new">Rediscover the web</A></b></font color=orange>
 

over_c

Distinguished
May 27, 2003
289
0
18,780
I'm not interested in debating the performance rating.

I just want to know how the AMD64 CPU/system that appears to be even or superior in every possible aspect over the Sempron CPU/system can lose any benchmark.
 

TheRod

Distinguished
Aug 2, 2002
2,031
0
19,780
Have you read <A HREF="http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=206576#206576" target="_new">my post</A>?

I found some reason why the new Sempron beat the A64 in a few benchmarks. But, generally the A64 is faster, so I don't bother much about those 2 lonely benchmarks. Who will buy a Sempron 3300+ based on these 2 benchmarks results? No one.

The K8 architecture is much less dependant on cache size than P4 architecture. It's not surprising at all that Sempron looks so good compared to A64, Sempron are much less crippled than Celeron.

There is no need to debate about the Sempron 3300+ performance since the CPU is not generally faster than is big brother!


-
GA-K8NF-9 / <b><font color=green>Athlon 64 3200+</font color=green> @ 3800+</b>
Infineon DDR400 (CL2.5) 2x512Megs
<font color=green>GeForce 6600GT 128Megs</font color=green>
<A HREF="http://www.getfirefox.com" target="_new">Firefox</A>
 

over_c

Distinguished
May 27, 2003
289
0
18,780
Yeah, I must have missed the meat of your post. I'm not really interested in debating which CPU is better, that is fairly obviously the AMD64. I just wanted to learn a little something more about CPU architecture. And I do believe the AMD64 use in the test was a Winchester, as they mention it has the dual channel memory interface.
 

sobelizard

Distinguished
Dec 31, 2002
418
0
18,780
<A HREF="http://www.voyeurweb.com" target="_new">Nerdy Stuff</A>

<b><i>Powered by <font color=blue>V</font color=blue><font color=purple>E</font color=purple><font color=red>R</font color=red><font color=purple>T</font color=purple><font color=blue>O</font color=blue></b>
Fueled by <b><font color=blue>CL-</font color=blue><font color=red>ONE</font color=red></b>
 

over_c

Distinguished
May 27, 2003
289
0
18,780
Thanks for that link. I opened it just as my supervisor happened to look over my shoulder. A little warning next time, perhaps?
 
LOL - I always put the mouse over the link to see what it is before I click! Saw Voyeur and decided it might not be prudent....

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red><P ID="edit"><FONT SIZE=-1><EM>Edited by rugger on 04/20/05 03:57 PM.</EM></FONT></P>
 

phial

Splendid
Oct 29, 2002
6,757
0
25,780
dont open links from work. i found this out the hard way...


or you coudl haev at least looked at the URL. its says VOYEURWEB, i mean common thats obviously something you should be careful of at work ;P

-------
Work sucks.
 

over_c

Distinguished
May 27, 2003
289
0
18,780
I guess I'm just used to the commonly used NSFW tag for things like that. Not on this forum, but on others. I just didn't expect to enounter a link like that outside of the "Other" board.
 

sobelizard

Distinguished
Dec 31, 2002
418
0
18,780
If your workplace truly didnt want you to visit that site, it would have been blocked.

Admit it, either you bookmarked it or memorized the url!

<b><i>Powered by <font color=blue>V</font color=blue><font color=purple>E</font color=purple><font color=red>R</font color=red><font color=purple>T</font color=purple><font color=blue>O</font color=blue></b>
Fueled by <b><font color=blue>CL-</font color=blue><font color=red>ONE</font color=red></b>