My opinion on the AMD PR BS

[xxsk8er101xx]

When i switched over to Intel, I personaly was fetup with AMD's lack of stable chipsets. It was pathetic. This was way before nforce 2 came out. nforce boards didn't appeal to me because they were all user boards. They were small, very little upgradability. So i said screw this and got intel. Abit IT7 just cought my eye. Had everything i wanted on a single board. And, most important, the only VIA on that board was for 2 USB ports.

Now AMD itself is getting rediculous. AMD finally fixes the chipset problems and the heat issue. Then they go and rip off consumers like this. It's a slap in your face. SO i don't blame you for being offended like this.

At the sametime it's just a product. If you don't want it don't buy it. AMD's management will have to explain to the CEO why the sales are soo low. You'd be amazed how DUMB mangment can be. I know engineers arn't that stupid so i'm betting the engineers made it a lower rating but managment said "no make it this and the worlds fastest cpu to increase sales".

thats what i think. It is rediculous.

"Bread makes me poop!" - Special Ed

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>

 

[sabbath1]

That's very true. The 3200+ is not the worlds fastest PC processor. Though it comes close, it's still a tiny bit slower than the P4 3.0. And AMD knows it. Obviously temporary increased sales matters more than reputation. Thus when reputation goes down, sales will too.


My system: AMD Athlon XP 3000+ CPU / TwinMOS 1Gb DDR400 / Soltek 75FRN-RL /
Sapphire Radeon 9800 Pro / Antec True Power 550W / Maxtor 80Gb ATA-133 / Hercules GTXP SC /
Samsung DVD / Lite-On CDRW

 

[TheRod]

There is one big problem in the processor industry... There should have a "Rating Standard".

We can't no more tells the TRUE performance of a CPU by checking is clock speed. We need something decent to compare CPU.

I wish to see an independant group to build a reference in benchmarking and rating for CPU. And then, you would buy CPUs with a "standard rating".

There should be 3 rating.
1. Gaming
2. Business / office
3. 3D / Video editing

Ta Dam!!! Make my dream come true!

--
Would you buy a GPS enabled soap bar?

 

[xxsk8er101xx]

ya wouldn't be that hard to implement either. However, it would cost money. so that will never happen.

"Bread makes me poop!" - Special Ed

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>

 

[sonoran]

>>>There is one big problem in the processor industry... There should have a "Rating Standard".

You mean like this: <A HREF="http://www.spec.org/" target="_new">http://www.spec.org/</A>?

And while I also like the idea of ratings based on popular uses/programs, those change all the time. That would make comparisons over time difficult and/or meaningless.

 

[eden]

Indeed. In the past we had huge debates over this, but the end result is, you simply can't. No matter what you try, the user will always use a program so obscure, but yet may not run in parallel performance to the overall.

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden.

 

[slvr_phoenix]

I think the big problem is that the entire PC affects performance, not just the CPU. At the very least the RAM, the mobo, and the BIOS settings all have a significant impact on benchmarks.

So you need to have a system that compares entire systems, not just CPUs. Frankly, the CPU's rating is just not the end-all of importance in the grand scheme of the PC's performance.

Which is what I've said in the past and what I'll say for years to come I'd imagine. If anything is done, what needs to be done is a US government requirement that <i>all</i> complete systems sold in the US come with a government-standardized set of benchmark results right on the outside of the packaging. It works for the food industry labelling nutritional values on pre-packaged food. It'd work just the same for OEMs labelling complete systems.

And if anyone wants to test my current sanity: "<i>Nobuo Uematsu - Melodies of Life</i>" rocks.

<font color=purple><pre><b>There are 10 types of people in this world: those who can understand binary and those who can't.</b></pre><p></font color=purple>

 

[imgod2u]

There is no measurement of "true" performance because there's no such thing as "true" performance. You can't have a "true" performance for games because games aren't all coded the same, you can't have one for business because there isn't even a standard set of programs that count as "business". Do you care about Word performance? Antivirus? Software makes the performance and this is especially true of different architectures. There is no "true" performance because every piece of software is different. The only consideration is what software is important to you, which makes processor performance subjective. There's no way to state definitively that "processor A is faster than processor B", merely "processor A is faster than processor B in software C".

"We are Microsoft, resistance is futile." - Bill Gates, 2015.

 

[JimStapleton]

I believe they are fixing this with the Opteron modle numbers which dont appear to even attempt to mimic Mhz ratings. The put them selves into a hole when they started these modle numbers, I think by the time they realized it, getting out with the Athlons would be more harm than good...

Anyway, the original numbers were designed to match the 2nd gen P4 core/bus combo, aren't we on the third or 4th? Initially they were QUITE generous to Intel with their model numbers, but the P4 got more efficient, and the modle number apparantly didn't.

Athlon XP 1600+, MSI K7T PRO2 RU (POS), 2x256 MB CRUCIAL PC2100 CL2.5 memory, Asus V6800 DDR Delux (GF 256) video card, 6.4GB+27GB WD HD, 40GB IBM HD (all 7200RPM). My computer is an acronym

 

[Spitfire_x86]

I like the Opteron model numbering, no place of MHz.

----------------
<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>

 

[cdpage]

Perhaps the modeling number system could work... however i think the problem is that it is Still modled around Mhz... perhaps it would be best to modle it after a benchmark or a combination of benchmarks. what ever the score is would be its modle number...

ie AMD 2345/120/666
PC 2100/200/420

it may be long... but think of the advantages.... at first you might think that most ppl might buy the AMD... but what if you a user that uses mostly on the middle benchmark...what ever that might be

Just a thought

dah, has any body seen my paints? :redface:

 

[sabbath1]

Despite that AMD may lose marketing power, they should drop their rating system. Right away.

My system: AMD Athlon XP 3000+ CPU / TwinMOS 1Gb DDR400 / Soltek 75FRN-RL /
Sapphire Radeon 9800 Pro / Antec True Power 550W / Maxtor 80Gb ATA-133 / Hercules GTXP SC /
Samsung DVD / Lite-On CDRW

 

[JimStapleton]

I admit, I'd like to se a model number in this format:

[Ghz * IPC * ISS]|[FSB]
Ghz = Gigahertz
IPC = Instructions per Clock
ISS = Size of instruction set
FSB = Front Side Bus

2 numbers, quite useful.

Athlon XP 1600+, MSI K7T PRO2 RU (POS), 2x256 MB CRUCIAL PC2100 CL2.5 memory, Asus V6800 DDR Delux (GF 256) video card, 6.4GB+27GB WD HD, 40GB IBM HD (all 7200RPM). My computer is an acronym

 

[Mephistopheles]

The problem is that IPC itself is not a constant... it may vary. The designed IPC is only a theoretical maximum, so...

Oh, and by the way, slvr_phoenix, Nobuo Uematsu rocks indeed! Nothing like ages-old final fantasy themes... nicely written stuff.

 

[slvr_phoenix]

There is no measurement of "true" performance because there's no such thing as "true" performance.

I completely agree. Just as there is no measure of "true" nutritional value because there's no such thing as "true" nutritional value. Vague idealistic concepts of a universal truth do not and should not hinder best efforts at providing more information so that consumers can make wiser decisions.

I can look at my 'Cup-a-Soup' and see that it has 50 calories and 10 calories from fat per 6 oz serving. I can look at my 'Instant Lunch' and see that it has 280 calories and 110 calories from fat per 14 oz serving. (Which makes 120 calories and 16 calories from fat per 6oz serving.)

From this I can deduce and make an educated decision that if I want to eat the least fattening meal, I should eat the 'Cup-a-Soup'. Don't you find it amazing that I can make such an educated decision even though there is no such thing as "true" nutritional value?

I'm not stating that the complete PC performance measurements will be perfect. The system of measurement will obviously have its faults and will clearly not directly translate to all software ever to run on that PC. However one <i>can</i> make educated inferrences based on these measurements <i>if</i> they are standardized. And further <i>if</i> I know how the measurement of performance for my particular software compares to these standardized measurements, then I can easily infer how well the PC will run my particular software with a fair (though not perfect) degree of accuracy.

That is life. Nothing is perfect. We don't just sit around and wait for the world to finally become perfect. We work with what we have and compensate by designing things in such a way that we can mitigate the fallibilities into acceptable margins.

There's no way to state definitively that "processor A is faster than processor B", merely "processor A is faster than processor B in software C".

Which I never once suggested. In fact this time I didn't even suggest doing this at the processor level, but at the entire PC level. It isn't about declairing processor A faster than processor B. It is about giving consumers the ability to easily decide if complete system A from Dell meets their needs better or worse than complete system B from Gateway or complete system C from Dell. Yes, to know that they <i>do</i> need to know their needs, but at least it will be a thousand times easier and more accurate than the means that we have available now.

I'm not suggesting a perfect solution, merely a vastly better solution than what we have at present.

<font color=purple><pre><b>There are 10 types of people in this world: those who can understand binary and those who can't.</b></pre><p></font color=purple>

 

[slvr_phoenix]

Oh, and by the way, slvr_phoenix, Nobuo Uematsu rocks indeed! Nothing like ages-old final fantasy themes... nicely written stuff.

Damn straight. I think I'm going to listen to my chocobo mix today.

<font color=purple><pre><b>There are 10 types of people in this world: those who can understand binary and those who can't.</b></pre><p></font color=purple>

 

[Gamerigs]

Has everyone forgotten how AMD denotes its XP designations? What they did is when they changes to the Athlon XP naming from the Athlon K7 the took a benchmark of a Athlon k7 1000Mhz processor and then measured how much faster the XP processor was in comparision. This then gave them the XPXXX rating that we all sort of regconise. This of course gave them an excellent marketing tool when comparing there processes with Intels. Athlon 3200XP looks much better than P4 3.02Gigs. However the Athlon processor probably isn't rated over 3.02Gig yet can preform as well if not better, so how does that work? Well it like saying a 6 cylinder engine can often out preform a 8 cylinder engine?

 

[Mephistopheles]

Athlon 3200XP looks much better than P4 3.02Gigs.

Hm... I wonder where gamerigs got that one from... ideas, anyone? Maybe he means that "3000"<"3200" or something... Besides, that thunderbird-rating excuse is just laughable. How many people out there do NOT use the XP____ to compare AMD processors with P4s? Even people who work with computers do that! And the PR ratings used to be a good marketing tool indeed, but they seem to be doing their job backwards nowadays...

 

[eden]

Are you an AMD PR executive or what?

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden.