What does the "3000+" stand for?

I know you guys have answered that question many times. And I honestly searched for an answer before I posted. But with a AthlonXP 3000+ according to information I gathered here runs @ ~2167Mhz.

I have heard two stories and dont know which one is true.
1.) 3000+ is a ploy by AMD to make users think they are getting a 3Ghz processor by giving it the 3000+ Model number (which I dont think they are doing that or atleast i hope not)

2.)It mean it "processes" its instruction set as fast as an Intel P4 3Ghz. Which if this is true.... can someone explain how AMD is capable of doing that? I always thought the faster Mhz you have the faster the machine will process information and spit it back out to the screen or whatever. Or show me a link where I can read it.

Thanks
Dan

Well, as you already guessed, the 3000+ stamp on a 2.1ghz chip is a performance rating.

There is a lot of controversy as to what AMD are comparing to, when they rate the chips, it could either be comparison to the first Willamette core P4s, or to the older AMD Thunderbird.... which goes to suggest that your 2.1ghz processor will process at the same speed as a Thunderbird running at 3ghz

I'm not 100% sure about this, I could well be wrong, but this is what I was lead to believe

-Blueacid

AMD MHz is not the same as Intel MHz because AMD CPU does more work per clock cycle.

The official explanation of the PR ratings is that they are not compared to Intel but to old AMD Thunderbirds, so 3000+ is how fast a hypothetical 3 GHz AMD Thunderbird would be.

The marketing part comes in when overclocking is concerned as all Thoroughbred B and Barton CPUs average 2.3 GHz which is in the 3000+s PR ratings wise.

If you were buying a Thoroughbred B, it's better to buy a $49 1700+ Thoroughbred B than a 2800+ Thoroughbred B costing hundreds of $ because they'll overlock similarily. Read below why.

The smart thing to do is to buy a 1700+ Thoroughbred B or a 2500+ Barton, then invest your $ in other components which will allow you to overclock the system higher than if you bought a 3000+ CPU, then skimped on other components.

c627627 said:

Overclockers exploit the manufacturing process knowing the goal of the manufacturer was to have all CPUs being capable of running as fast as the line's fastest processor plus additional headroom. Only those CPUs that pass rigorous tests at default voltages, are given labels toward the end of the line. Others are labeled not just according to tests, but according to marketing plans.

So as you can see:
http://www.pbase.com/image/17079307/original

Thoroughbred B's scale to 2800+.

We noticed that the greatest overclockers were 1700+ and 2100+ T-Bred B's capable of reaching those end of the line speeds of 2800+ "plus additional headroom."

That's why it's best to buy those CPUs and not higher labeled CPUs since $ is another factor in overclocking.

...and that's why 1700+ T-Bred B should scale not too far below any 2x00+ T-Bred B using the same equipment.

Click to expand...

Consider this anology of an engine. AMD can create X amount of horspower at 3000rpm. Intel can create that same amount of horsepower, but only after revving up to 4500rpm. Same power, different speed.

That's really rough but I think it works.

brennan77 said:

Consider this anology of an engine. AMD can create X amount of horspower at 3000rpm. Intel can create that same amount of horsepower, but only after revving up to 4500rpm. Same power, different speed.

That's really rough but I think it works.

Click to expand...

I use that anology all the time, however the person must be able to understand the concept of RPM and HP too

Try this one on for size. Intel uses a 4 cylinder engine and a blower to get 3000. AMD uses a 6 cylinder engine and fuel injection to get 3000.

Ok

Alrighty.... that seems to clear up alot of stuff conserning the numbering system. I think explaining to my father in terms of RPM and Horsepower will be the best seeing how hes owns his own auto repair business.... but the technical definations are greate for me so I can know what really going on.
Thanks everyone!

Happy motoring!

This is an interesting subject... It was explained by OC.com way back at the end of '01, but the majority of the public still does not know/understand the truth on the matter, but thats just because most people couldn't give a rats ***... but we do care, so it's necessary to understand it correctly.

Spuppy21 said:

2.)It mean it "processes" its instruction set as fast as an Intel P4 3Ghz. Which if this is true.... can someone explain how AMD is capable of doing that? I always thought the faster Mhz you have the faster the machine will process information and spit it back out to the screen or whatever. Or show me a link where I can read it.

Click to expand...

You are on the right track here. I will link you to an article by overclockers.com at the end of my post so that you have a solid reference to know the true facts from - not the fanboyism that sometimes floats around and people hear about (think -> your first option).

To simplify things, there are two choices a chipmaker can make. They can:

• Do more work less often or
• do less work more often.

If they choose A, then their chip won't run at as high of a frequency. If they choose B, then their chip will run at a higher frequency.

Herein lies a game of give and take... If you want to increase the frequency it makes for a smaller workload per Mhz, while if you want to increase the workload per Mhz it makes for a lower frequency. So there is an inverse relationship between workload and frequency, and this means that an increase in one can make up for a decrease in the other so to speak. So that's that.

Actually, just because I like you, I'm gonna give you a few free bonus links along with the original link I planned to give you. Before the links, I will give the most important quote from Ed:

Ed Stroligo on the institution of the PR rating system
An Admission of Desperation

This is a desperate move by AMD.

Not desperate as in "we're desperate because our processors can't compete" but "we're desperate because we have no chance convincing those $%*@#!! dimwits out there that our 1.4GHz Athlon is faster than a 1.7Ghz PIV (or equivalent)."

I think this is an area where it's safe to say AMD should have a far better idea than we do as to how hopeless explaining that to the general public will be.

Click to expand...

This is the article I originally intended to link you to:
"PR Or BS?"
Ed Stroligo - 8/29/01

Here are your other freebies on the subject, they paint an interesting picture:

"An Enema for the Anal-Retentive"
Ed Stroligo - 9/1/01

"No Comprendo"
Ed Stroligo - 9/7/01

"A Matter of Faith"
Ed Stroligo - 9/8/01

"A Little Math"
Ed Stroligo - 9/10/01