THG still claims to AMD ?

[vgrigor]

THG often claims to AMD - to reduce their last processors rating...
May be somw true there is at this, but -
does they selves did any step to cooperate with other testers? Did they respond to many different opininons of other people here, around, and collegues?
To show sample - how to do claim -in right way.
No - They did not.
(According to their relations of revievs to people posts
with questions and results of it).

Is is correct?
Or I something misunderstood?

 

[shadus]

I'm not sure anyone can quite understand that one. It looks like it went through babbelfish.

The high end AMD processors are overrated by a pretty significant clip according to the reviews here and most reputable places on the net. The older AMD processors were generally underrated.

Shadus

 

[vgrigor]

Yes, some not placement of some source information -
read it here:

http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=136229#136229

 

[unoc]

AMD introduced their P rating to make a comparison against the P3. Now, the P4 is more efficient and the AMD P rating is using units of measurement different from that used by Intel which uses the frequency at which the cpu operates.
I believe that the core frequency is much more an absolute value than a mere comparison with a processor not in production. It is as I would compare a Ferrari with an AMD car which claims to run not at 300 km/h but at 600 times an ant running.
AMD understood that a lot of people which buy computer in supermarket are really struck seeing a computer with an AMD XP CPU working at 2800 + (what "+" means? maybe also more than 2800 MHz ?)

______________________

 

[vgrigor]

The real comparison - that is most often already used,
and understood by wide of people -
Mhz -"made by Intel".

 

[Spitfire_x86]

AXP original PR was designed to compete with Nortwood "A". They were underrated for Willamette (1800+ was able to beat 2.0 GHz Willy). AMD revised their PR system for Northwood "B", it was good like initial PR (T-bred 2800+ beats 2.8 "B" P4). AMD finally revised PR for Bartons, that was really bad. AXP "Barton" 2800+ is most times slower than "T-bread" 2800+, let alone 2.8 GHz P4 "C".

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A></b>

 

[Twitch]

Actually, technically, the PR rating was supposed to relate to an Athlon Thunderbird core...an 1800+ was theoretically as fast as a Thunderbird at 1.8 ghz...

However, the problem is that in actual practice, the PR rating doesn't to relate to anything except for AMD's arbitrary need for a high model number. Someone explain to me how AMD can justify adding 33 MHZ to their FSB speed and 33MHZ to their core speed, and then upping the PR rating by 200 mhz!!! So obviously, PR rating relates to nothing.



<font color=green>The Netherlands is where you go when you're too good for heaven.</font color=green> :tongue:

 

[Spitfire_x86]

AMD always told in their AXP website that AXP ****+ is faster than P4 ****. This PR system compares with P4, not T-bird

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A></b>

 

[Twitch]

A lot of people think that, but it's not, technically what the model number is supposed to represent.<A HREF="http://www.geek.com/news/geeknews/2001oct/bch20011008008225.htm" target="_new">Link</A>

It's not worth arguing over anyway, since the PR rating obviously deosn't mean anything.

<font color=green>The Netherlands is where you go when you're too good for heaven.</font color=green> :tongue:


<P ID="edit"><FONT SIZE=-1><EM>Edited by Twitch on 05/28/03 11:38 PM.</EM></FONT></P>

 

[SMellery]

Waht is even more interesting still is that the new AMD Barton 2500+ benchmarks faster than the T-Bred B 2600+, but all the other bartons bechmark lower than their t-bred counterparts. When will the mayhem stop!?

 

[Turk]

+200mhz rating because of a bigger cache?

Turk

 

[Groveling_Wyrm]

*ahem ahem* politically correct here....

its not 200 mhz....its 200 PR points....

With Intel's method, you can't make something and just give it any name you want, as it would "appear" that AMD did. You must actually stick to a set standard (the Mhz standard).

Unfortunately, AMD was losing the battle by using a standard that was standardized by Intel, so they had to find their own. Now they need to create a standard that is credible to all.

<font color=blue> Ok, so you have to put your "2 cents" in, but its value is only "A penny's worth". Who gets that extra penny? </font color=blue>

 

[confoundicator]

<A HREF="http://www.anandtech.com/IT/showdoc.html?i=1817" target="_new">http://www.anandtech.com/IT/showdoc.html?i=1817</A>
Personally I'm glad the new Opterons are using pretty much arbitrary model numbers. The Opteron 244 (<--notice meaningless model number) is a 1.8Ghz chip, but competes with and sometimes outperforms the 2.8Ghz Xeons. Guess we'll have to wait and see if the desktop versions will have a different numbering system than the business versions.

I'm just a big, geeky nerd.

 

[Twitch]

You're right about that. And notice the new model numbers are truly arbitrary in relation to Intel's processor speeds. I don't have any problem with that at all. However, the Athlon XPs correlate with Intel's speeds which is the problem.



<font color=green>The Netherlands is where you go when you're too good for heaven.</font color=green> :tongue:

 

[endyen]

That is correct twitch. Is there anyone out there that thinks an xp1400 is only as good as a willy 1.4 gig?
A number of benchmarks show a fairly linear progression for all xp procs, including the 3200. Most of us are disapointed because intel made such big gains with the 200 fsb chips, not to worry ATHLON64 is coming !!!

 

[eden]

Um that is a pretty bold statement, both for the disappointed part and the Athlon 64.

First, the gain from 200MHZ FSB and memory is not this huge, even with PAT. It could've been much higher had it properly been used. As of now, only 5GB out of the theoretical 6.4GB of per-second bandwidth is used. And that is with PAT. Take PAT off and you remove about 400MB/sec. Neither chipsets are anywhere near the limit and are definitely unoptimized. Comparatively, AMD's latest systems have been about 200MB/sec shy of the limit, and the K8 memory controller has about a 96% efficiency, leading us to a very near theoretical usage limit.
My goal is not here to praise AMD, but to tell you that you are speaking for yourself NOT for everyone. I was disappointed with the LOW 200MHZ FSB gain, and however am happy that Intel is winning, because it only proves AMD is not trying anymore.
As well, I am not putting my hopes on Athlon 64 anymore. We know nothing of the improvements potentially done during this period, and therefore I will just go with the flow.
The Prescott processor is nearby, and 3.2GHZ Pentium 4s are nearing by. Hyperthreaded P4s are ranging from 2.4 to 3.06GHZ. All the more reasons to go Intel right now.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue: