How can a 1.4 T-bird OC'd to 1.6Ghz be better than Athlon XP 1600+?

[Plankton]

Athlon XP Model Numbering System

Starting with the Athlon XP processor launch, AMD will be introducing their new model numbering system as a bridge to their plans for a unified number system in 2002.

The new AMD Athlon processors will be sold on the basis of a relative performance scale. These models numbers will pertain and convey a performance level relative to other AMD processors. As of now, the Athlon XP processor model numbering will be based on the Thunderbird core Athlons. Below you will see the chart of AMD Model numbers and their actual operating frequencies.





The new Athlon XP 1800+ processor will actually be running at a speed of 1.53 GHz, not 1.8 GHz as the name might suggest. Similarly, the 1700+ runs at 1.47 GHz, the 1600+ runs at 1.4 GHz and the 1500+ operates at 1.33 GHz. AMD’s new scale compares their new processor to their Thunderbird. This means that the Athlon XP 1800+, which runs at a clock speed of 1.53 GHz, performs nearly the same as an original Thunderbird core Athlon processor would at 1.8 GHz. Along the same lines the Athlon XP 1700+ running at 1.47 GHz performs like a Thunderbird would at 1.7 GHz, and so on and so on


Hi, I don't know if this will help but its right from www.amdmb.com i just cut a small section out but they have a really good article all about the xp and what the numbers mean and what the numbers are based on.

 

[c627627]

AMD’s new scale compares their new processor to their Thunderbird. This means that the Athlon XP 1800+, which runs at a clock speed of 1.53 GHz, performs nearly the same as an original Thunderbird core Athlon processor would at 1.8 GHz. Along the same lines the Athlon XP 1700+ running at 1.47 GHz performs like a Thunderbird would at 1.7 GHz, and so on and so on

Thanks Plankton, I knew I read that somewhere. So at least some of the places out there say that Athlon XP ratings are an indication of Thunderbirds performance in megahertz. But I also think Silver's work is valid and that in reality if a Thunderbird can run stable at say 1600MHz, it would outperform XP 1600+ at stock speed of 1400MHz.

Ofcourse, cooling is a separate issue and XP is superior there, so it would come ahead. Here's a question for experienced overclockers: Let's say you have two systems whose only difference is AMD Thunderbird 1.4GHz and AMD XP 1600+ (1.4Ghz), How high percent wise do you think each of these could overclock on average presuming you can only use retail cooling?

 

[Silver]

Now that one is a silly one. My t-bird will cook long before my xp1600 will. Heck with a ocz gladiator it will 1.45 and that is it. Now with water there is about another 300mhz in it.

 

[Silver]

We used H-Oda's popular WCPUID to see what it can tell us about the inner workings of AMD's new flagship processor. As most of you already know, the 2100+ moniker is not used to denote actual clock speed but rather the processor's relative performance rating, when compared to other x86 CPUs in its class.

http://www.hothardware.com/hh_files/CCAM/amd_axp_2100(2).shtml

This contentious issue's inherit aspects of beginner deception is now largely overlooked purely because the 1600+ would indeed best the performance of the competitions 1.6GHz CPU. In fact, as AMD have proved time and time again in a variety of vastly different benchmarks, it not only (and forgive me for paraphrasing myself) 'bests it', it smashes through the performance levels set by competitor's products.

http://www.3dvelocity.com/reviews/amd_athlon/xp2100.htm

 

[Silver]

Ahh, this did not come from a regular old site, this came from AMD. Yup that OF COURSE is what amd is saying. Everyone knows the xps are compared to The T-bird 1.4. Hmmmm, check out AMD for the real low down.

 

[Silver]

And the testing goes on and on and on. Private research testing paid for by amd, etc, etc, etc......need to do a little more research and do not go to the trough, go to the source. Take Care.

 

[Kendan]

BTW where does one come up with the info that a T-bird would have to run this fast to match a xp1600? Just curious as for the life of me (other than in forums) I just can not find any data to concur with this.

I read it from AMD but I have not been able to find it. It stated that AMD used the TB 1.4 as a baseline in determining the XP model # system. Of course they are going to compare their processor to the P-4, but that does not mean they are using it to determine their model # system. If that were true that they used the P-4 then a 1600+ XP would be equall to a P-4 1600 which we all no is not the case. The XP 1600+ will waste the P-4 1600. And if you go to this page . it states that the model # system is relative to the current Athlon. It does not say the P-4.

 

[Silver]

Yes and they specifically refer it to the "Athlon XP Processor". Having directed one to the amd white papers I am sure that you have read them and know after reading them that the xp specification is irrelevant in regards to the Athlon. If not then you have not read the document.

 

[Silver]

This alone should have sufficed. Read the document. Rather simple math shows you that AMD themselves show a what 13% advantage on the 1.4...hmmm 1400x1.13=1582....obviously that ain't a 1.4...No where do they state that a T-bird 1.4 would have to run at anything nor do they state that the xp series represents what a 1.4 would have to run at to match an xp. They do however state that the 1.4 T-bird was used as the baseline. AS a matter of fact after reading the article one has every reason (as indicated by AMD) that the intent was to indicate the "feel" of a comparable Intel. However I do appreciate your thinking as my T-bird gets faster by the day....It is a 1400 running at 1.73 which pr by sisoft at 2300. This is really good as at a pr of 2300 that means that my t-bird is running at what 2800, maybe 3000. That I believe is an erroneous statement on the amdmb board and should be easily quotable by amd. After all if it is your chip then why not state that? It is not stated on the site as they really do not want to make the statement against Intel...why, potential liability. Ask yourself another question....why does'nt sisoft have a T-bird comparison in the Sisot CPU benchmarks. After all it is the top of the Athlon line and the baseline for the xp's. Might have something to do with the relatively small increase if the software enhancements are not figured in? Come on... AMDMB stands for AMD mother board....(as in possibly Socket A?)......Somehow people see AMD at the front of that and figure well that is an official AMD site....wrong motherboards, motherboards....

 

[Silver]

And that is exactly what the results above show. The statement is true however your implication that it is a reference to the Athlon xp being named according to the Athlon 1.4 is at the least misleading to anyone reading this thread. Below is your reference. You really ought to read the whole thing.

» Technical Information » AMD Athlon™ XP Processor Benchmarking and Model Numbering Methodology


AMD Athlon™ XP Processor Benchmarking and Model Numbering Methodology
QuantiSpeed™ Architecture

Understanding Processor Performance

AMD Athlon™ XP Processor Benchmarking and Model Numbering Methodology


Desktop processors based on the "Palomino" core will be marketed as the AMD Athlon™ XP processor. As a way of communicating the performance improvements of the new AMD Athlon™ XP processor relative to the performance of the currently available AMD Athlon™ processor, AMD has developed a model numbering convention.

Go look around and you will find that AMD is using a very arbitrary means of deciding on the performance of the xp series. Not really based on any static industry accepted standard. Sure works out nice relative to the Athlon to move up one on each new chip...works out really nice when those numbers just happen to match identically with the competitors actual cpu speed. It must be me as how in the world can people really think this.

 

[Silver]

If that were true that they used the P-4 then a 1600+ XP would be equall to a P-4 1600 which we all no is not the case.

Take a good look at the 1600+ XP that you typed up there and tell me please what the "+" means to you.

Did you read this whole forum....I did not make that stuff up....those are from AMD not the site AMDMB (motherboards).

 

[Silver]

The new Athlon XP 1800+ processor will actually be running at a speed of 1.53 GHz, not 1.8 GHz as the name might suggest. Similarly, the 1700+ runs at 1.47 GHz, the 1600+ runs at 1.4 GHz and the 1500+ operates at 1.33 GHz. AMD’s new scale compares their new processor to their Thunderbird. This means that the Athlon XP 1800+, which runs at a clock speed of 1.53 GHz, performs nearly the same as an original Thunderbird core Athlon processor would at 1.8 GHz. Along the same lines the Athlon XP 1700+ running at 1.47 GHz performs like a Thunderbird would at 1.7 GHz, and so on and so on


This statement (not from AMD) does appear to be erroneous given AMD's (the cpu maker) above referenced graph.

 

[Silver]

I read it from AMD but I have not been able to find it. It stated that AMD used the TB 1.4 as a baseline in determining the XP model # system. Of course they are going to compare their processor to the P-4, but that does not mean they are using it to determine their model # system. If that were true that they used the P-4 then a 1600+ XP would be equall to a P-4 1600 which we all no is not the case. The XP 1600+ will waste the P-4 1600. And if you go to this page . it states that the model # system is relative to the current Athlon. It does not say the P-4.

Generally what you have said here is true.

Yes, as one can see from the graphs the 1.4 (1400) was used to number the series. I.e. first 1400 (1.4) then 1500, 1600, 1700, etc. This is NOT a reference to performance in comparison of it to the T-bird 1.4 as one can easily deduce from AMD's submitted data.

This just proved to be a really good point to enter into the arena with matching Intel processors. No where is it stated specifically the the increments between the number has to be 66 mhz and tomorrow they could (bad marketing financially) if forced to name the next one to come an xp2200 (xp2100 is out) and make it a 200mhz increase) of course one can see that this has no bearing on the performance of it in comparison to a T-bird 1.4, it would just be the next in the series relative to the T-bird 1400.

 

[Silver]

I read it from AMD but I have not been able to find it. It stated that AMD used the TB 1.4 as a baseline in determining the XP model # system. Of course they are going to compare their processor to the P-4, but that does not mean they are using it to determine their model # system. If that were true that they used the P-4 then a 1600+ XP would be equall to a P-4 1600 which we all no is not the case. The XP 1600+ will waste the P-4 1600. And if you go to this page . it states that the model # system is relative to the current Athlon. It does not say the P-4.

And in leaving (should you read the document that you referenced) then you will as they state learn more about the xp numbering. First you will see the chart shown here in this thread then you can look at all the comparisons of the xpxxx with its' comparable intel in mhz being tested side by side and when you are done looking at these you will have learned more about the xp number sequence and it's performance against the same intel counter part at the same number in mhz.

lastly, xp stands for what? Why would one call a slower processor "extreme performance"? Might'in it be because though it is slower it performs as well as ......a T-bird 1.4?.....think...

xp = extreme performance, xxxx is next in series relative to the Athlon 1.4, + is a little better than the competitors comparable in mhz. and thus we have the xpxxx+ (xp1500+, xp1600+, etc.)

Using this logic, I bet one will find the xp2100 be tested between an intel 2000 and and intel 2200. Check around and see if this might be true....as a matter of fact check out AMD (as in AMD corporation) and see if they might have this test data available to the public, as in good marketing. Bet they do.

Do think about this, according to the graph if the T-bird 1.4 and the xp are run in a non-windows environment, how much advantage does it have? Really does not have a whole lot does it. Also note (when reading the information at AMD) that the testing between the AMDs and Intels were in many cases based on "feel".This should not have to be arbitrated as it would be complete ignorance to state/think that the AMD beats its comparable competitors (wonder whom) processor out in all areas. I do not think that you will find many if any reviews that will tell you that, much less AMD (again this is indicated by AMD in their own test data and available to the public at their site).

 

[Silver]

Athlon XP Model Numbering System

Hi, I don't know if this will help but its right from www.amdmb.com i just cut a small section out but they have a really good article all about the xp and what the numbers mean and what the numbers are based on.

Had to do a search on old reviews of xp1800 and technology review at the site you linked to here. Please note the following from the article. Quote from the article. Much of what they state is taken from AMDs White Paper.

"Those people see the name Intel, assume it’s the best, and just buy it. AMD’s goal is to make them think twice before making that snap-decision, and then realize the Athlon XP is the way to go.
How is AMD planning to do this? AMD is starting and initiating the development of new and fully encompassing measure of processor performance that customers and end-users can trust. Their goal is to not just include the AMD line of CPUs, but all x86 architecture processor in this new, industry-developed standard, including Cyrix and Intel. AMD will also be working with the lead industry players in the development of this standard including OEMs, channels, software developers and their many infrastructure partners. "


Finally found your quote. Unfortunately this quote was AMDMB's and not AMD's. It is not consistant with AMD's White Paper or other info available on the site. Furthermore they proved their statement by running side by side tests comparing the xp1800 (1.53) with the T-bird at 1.4. That the 1.53 beat the 1.4 proved indeed that the 1.4 would have to run at 1.8 to match it. Let's see, According to the graph the xp1800 runs about what 13% faster and geuss what it beat the T-bird by on average approximately 13%. hmmmm Author of the article came up with that on his own as AMD does not state that anywhere on the site.

In fact AMD makes many references to the disparage between the two (intel) and the need for a better naming of their chips so that the consumer better understands this. The testing gives virtually no data to compare with the 1.4 to prove their stance but uses the intel comparable number in mhz pretty much throughout the article. The above graph is about the only reference to a 1.4 T-bird.

 

[00chit]

Q: What’s the meaning of the “XP” in the AMD Athlon™ XP processor name?

A: The AMD Athlon™ XP processor is the name of the newest desktop processor from AMD. The “XP” modifier is designed to convey the extreme performance AMD Athlon XP processors deliver for the Microsoft® Windows® XP operating system.

ooohhhh big clue as to the software driven performance.

Ok, so now for my "bonehead" question...
If you aren't running WinXP then you aren't getting the full potential out of the XP cpu?

 

[Silver]

Ok, so now for my "bonehead" question...
If you aren't running WinXP then you aren't getting the full potential out of the XP cpu?

Sad but true....this I'm sure explains why my xp is parked in the bedroom. I may however put it in and give the xp1600 another go when the pelt finally gets here. Might be kind of fun to see what a 226 will do to a xp1600. donny was getting 1.7Ghz on his with an 80 watt and could not push it any harder as it overwhelmed the pelt.

 

[Plankton]

Hello...

I found another nice little article about the naming of the xp chips from www.sharkeyextreme.com Which had the following i thought this is one of the better ones cause you see it from both ends..

Remember the PR Rating?


If you were following the industry a few years ago when the legacy Pentium was king of the hill, you may recall the PR rating. It was a product naming method used by companies like Cyrix to denote the relative performance of their processors to what Intel was offering. The logic was that customers needed to be aware that even though the clock speed of a Cyrix processor was lower than a competing Intel processor, its performance might be equal or higher. As such the "Cyrix 6x86 P166+" and similarly named processors hit the market to the lukewarm reception of buyers who began to worry that they were being taken advantage of. It seemed, at the time, while they were correct in their assertion that their processors performed better than identically-clocked Intel processors, exactly where the performance feel in comparison to a competing Pentium processor was being determined by marketing reps.

It was a bold experiment, and it did not end well for Cyrix, which is now a not-often (nor loudly) heard from subsidiary of VIA Technologies. AMD, which also tried jumping on the PR bandwagon back at the same time, obviously walked away from the ordeal with less war wounds. And they've gone on to produce some the most impressively-performing processors on the market, still proving their processors are faster than similarly-clocked Intel processors, but doing it without the PR rating... until now.

For here we are, and here they at AMD are, once again struggling with the issue of wanting to demonstrate to Joe Consumer that their 1.4GHz processor is faster than Intel's 1.4GHz processor; indeed, faster than Intel's 1.7GHz processor in most benchmarks and real-world applications. And how have they chosen to do it? Well, they're not calling it a 'PR' rating this time around... in fact, their explanation of the new model-naming scheme states that it is not meant to compare performance to similarly-named Intel processors (rather to legacy Athlon speeds at the same clock speeds). But if it walks like a duck, and talks like a duck, and... well you know how the saying goes.

So What's the Naming Convention This Time Around?


Well in some ways, AMD has created a fairer naming convention than the PR ratings of yore. But unfortunately, they've also come up with one that is going to do more to confuse the average consumer than to educate him/her. At first glance, one would likely assume that a name of "Athlon XP 1800+" would mean that it claims superior performance to a 1.8GHz Intel P4. After all, that's what it meant the last time we saw these kinds of processor names. But AMD had previously made it clear that their new model naming scheme was meant to compare the performance of an Athlon XP processor not to an Intel processor, but to a legacy AMD Athlon. In other words, an AMD Athlon XP 1500+, while clocked at 1.33GHz, will deliver equivalent performance to what the Athlon Thunderbird would have had at 1.5GHz. From there, you should then know that since a 1.4GHz AMD Athlon Thunderbird processor was already matching the performance of a 1.7GHz Intel P4, now it's being matched by a 1.33GHz processor... still with me?

But then, in contrast to this seeming clarification, a FAQ on AMD's web site says that "The AMD Athlon XP 1900+ will outperform an Intel Pentium® 4 processor operating at 1.9GHz on a broad array of end-user applications," leaving us with that older and more familiar scheme that PR ratings used. Beyond that problem is also the fact that if the model name is meant to demonstrate its superiority to the Intel architecture, then what will the average consumer take away from the fact that even with the new naming scheme, they're still seemingly behind Intel's fastest processors? Basically the average consumer will see them as saying "our top-of-the-line AMD Athlon XP 1900+ is as fast, or faster, than Intel's 1.9GHz," when Intel has had a 2GHz processor available for months and is expected to release a 2.2GHz processor in the near future.

So the average consumer, who one must assume AMD is targeting with this new model naming convention since enthusiasts are already well aware of the great performance of AMD Athlon processors and would expect nothing less from AMD Athlon XP processors, is likely going to end up with more misconceptions about the whole issue than he/she did before the new model naming scheme. Not only will such consumers be unsure as to whether they're buying 1.53GHz processors or 1.8GHz ones when they choose the AMD Athlon XP 1800+; but, even more importantly, many are going to see this as a sign that AMD is a follower, not a leader. Intel, after all, set the standard by which AMD will appear to measuring their processors (though as I earlier noted, the multiple explanations and clarifications on this matter have left that unclear) and still behind in the race despite the new names.



I know this is just another "erroneous" post...LOL.

 

[Malakai]

Silver, good comments(nuthin i didnt already know though )
AMD is real good at putting up how they come up with results for their testing, and the actual benchies they use.
if your an AMD lover, u should browse their site often, they put up interesting articles all the time. I go there all the time cause I want to work their R&D one day.
-Malakai

 

[Silver]

Plankton, first "Welcome to the Forums". The post is in fact pretty good.