AMD's PR system is now totally useless!

[HardWareXpert]

<<

<<

<< My claim still stands that the XP rating system will have to be changed in the not to distant future... And I think this is a bad thing. >>



Are they being deceptive, YES. If the public thinks its a 2000 mhz processor, and it is not, THAT is deception. >>




Are you mad?, I knew full well the Mhz of my AthlonXP 1700+(1.466Ghz) but you've totally missed the point.

Your still thinking in MHz, it's not a 2Ghz CPU(AthlonXP 2000+) but the performance it achieves is equivalent to that of a 2Ghz P4, And benchmarks PROVE this, so what is the customer losing, money?, no, performance?, no, Bragging rights?, yes maybe.

 

[RedShirt]

<<

<<

<<

<< My claim still stands that the XP rating system will have to be changed in the not to distant future... And I think this is a bad thing. >>



Are they being deceptive, YES. If the public thinks its a 2000 mhz processor, and it is not, THAT is deception. >>




Are you mad?, I knew full well the Mhz of my AthlonXP 1700+(1.466Ghz) but you've totally missed the point.

Your still thinking in MHz, it's not a 2Ghz CPU(AthlonXP 2000+) but the performance it achieves is equivalent to that of a 2Ghz P4, And benchmarks PROVE this, so what is the customer losing, money?, no, performance?, no, Bragging rights?, yes maybe. >>

>>



The average joe blow thinks he's getting himself a 2000 Mhz Machine. I am not mad, not at all. Heck, I think AMD has the best processors for the money out right now. I am the happy owner of two rigs that are AMD powered.

Why can't AMD just say AMD Athlon XP 1.67 GHz (Comparable to a PIV xxxx MHz). Their scale is going to be messed up. Right now it is ok, but with these new Pentium IV chips with the higher bus speeds, if they maintain the same scale, it will not be accurate, which is what I have been saying since post one of this thread.

 

[LikeLinus]

<<The AthlonXP 1800+ in benchmmarks is able to out perform the P4 1.8Ghz, HOW SIMPLE DO YOU WANT IT?

AMD only went from 1.43Ghz(1600+) to 1.47Ghz(1700+) because that jump was equivalent to a T-Bird 100Mhz jump. >>

LOL That has absolutely NOTHING To do with the stuff you were talking about. You are stating that a p4 using RDRAM loses to an AMD XP using SDRAM.

PROVE THESE BS STATEMENTS YOU ARE MAKING. LISTED BELOW

1. <<RDRAM can but against your standard AthlonXP/SDRAM it's not very good at all.>>
2. <<Fact is not many apps or game take advantage of RDRAM's huge bandwith> LOL What a joke.
3. << It's been proved that the P4's quad pumped bus and it's huge bandwith is pritty useless>>
4. <<AMD know that huge bandwith figures mean nothing and in reality>>

Dude I hope you don't believe the BS spewing from your mouth. I also hope to your parents are wearing boots right now.

How else are AMD supposed to tell the customer? Well ever heard of advertising and marketing? Commericals? Ads in magainzes? Having PCWorld do testing? They don't have to create some magical rating scale to lie to customers and misguide them in buying a product. No one is saying it isnt as fast. Still you miss the point.

AMD should be trying to advocate for a real PR system. THESE are the things they should be doing.

<Athlon 1.2Ghz/RDRAM> When did AMD start using RDRAM? lol. Also you do realize that the P4 losing to the AMD has really nothing to do with the ram it uses. Its the pipelines that cause the P4 to slow down. Intel lengthened them so they could scale higher becaues AMD started a mhz war. Now AMD can't keep up, and they wanna play this whole "oh mhz dont matter". Ignore the fact that AMD was the one with commercials saying that they were the first with a 1ghz chip and all the ad's. lol

You also realize that most industry people consider RDRAM superior to DDR? But you can still sit around and tell everyone that it's "useless". hehe

I nominate you for noob of the year.

 

[HardWareXpert]

Ok, Intel are at 2.4Ghz(2.64Ghz filtering through soon) now and AMD are at 1.76Ghz, people would think that the AthlonXP@1.76Ghz is dreadfully slow compared to Intel's offering @2.4Ghz but the is just not true.

THe best way the convince people that the AthlonXP is more effient at the lower clock speeds is the PR system, never mind whether they think it's a AthlonXP 2Ghz or not.

The main thing is there getting a CPU that performs solidly againest Intels Higher Mhz P4's even at a much lower clock speed AND PRICE I MAY ADD. If there any deceiving going on it's Intel saying BIG Mhz equels superoiur performance and selling SDRAM version of the P4.

 

[HardWareXpert]

<< <<The AthlonXP 1800+ in benchmmarks is able to out perform the P4 1.8Ghz, HOW SIMPLE DO YOU WANT IT?

AMD only went from 1.43Ghz(1600+) to 1.47Ghz(1700+) because that jump was equivalent to a T-Bird 100Mhz jump. >>

LOL That has absolutely NOTHING To do with the stuff you were talking about. You are stating that a p4 using RDRAM loses to an AMD XP using SDRAM.

PROVE THESE BS STATEMENTS YOU ARE MAKING. LISTED BELOW

1. <<RDRAM can but against your standard AthlonXP/SDRAM it's not very good at all.>>
2. <<Fact is not many apps or game take advantage of RDRAM's huge bandwith> LOL What a joke.
3. << It's been proved that the P4's quad pumped bus and it's huge bandwith is pritty useless>>
4. <<AMD know that huge bandwith figures mean nothing and in reality>>

Dude I hope you don't believe the BS spewing from your mouth. I also hope to your parents are wearing boots right now.

How else are AMD supposed to tell the customer? Well ever heard of advertising and marketing? Commericals? Ads in magainzes? Having PCWorld do testing? They don't have to create some magical rating scale to lie to customers and misguide them in buying a product. No one is saying it isnt as fast. Still you miss the point.

AMD should be trying to advocate for a real PR system. THESE are the things they should be doing.

<Athlon 1.2Ghz/RDRAM> When did AMD start using RDRAM? lol. Also you do realize that the P4 losing to the AMD has really nothing to do with the ram it uses. Its the pipelines that cause the P4 to slow down. Intel lengthened them so they could scale higher becaues AMD started a mhz war. Now AMD can't keep up, and they wanna play this whole "oh mhz dont matter". Ignore the fact that AMD was the one with commercials saying that they were the first with a 1ghz chip and all the ad's. lol

You also realize that most industry people consider RDRAM superior to DDR? But you can still sit around and tell everyone that it's "useless". hehe

I nominate you for noob of the year. >>






<< Athlon 1.2Ghz/RDRAM >>



Thats a typo stupid, you should have pick that up yourself



<< RDRAM can but against your standard AthlonXP/SDRAM it's not very good at all.>> >>



See my link above, games that are CPU and FPU intensive the Athlon wastes the P4, plan and simple



<< <<Fact is not many apps or game take advantage of RDRAM's huge bandwith> LOL What a joke. >>



Ask yourself why in bechmarks with Quake3 engine(in Anandtech own reviews) the P4/RDRAM is able to perform better than AthlonXP nearly all the time(forgeting that P4 is 100's of Mhz ahead)



<< It's been proved that the P4's quad pumped bus and it's huge bandwith is pritty useless >>



See same reviews on P4



<< AMD know that huge bandwith figures mean nothing and in reality >>



AMD's own FSB(266Mhz DDR) proves you dont need heavy bandwith for todays and tomorrows games, apps to perform on par with a 400Mhz QDR FSB of the P4 and RDRAM.

 

[LikeLinus]

Failed to answer a single one of my questions. Why am i not surprise.

Also, for the intelligent impaired (Yes that means you), we are not saying that AMD doesnt need a PR system. Jesus how many times do we have to state this? Do you have some kind of ADD or memory problems? There needs to be a non-bias PR system.

Done with you...it's boring to have to say the same stuff over and over to get it through your head

 

[HardWareXpert]

Seems you have not got the intelligence to read Anandtech's own reviews and benchmarks for yourself, as i've said counless times the PR system is not biased as even the benchmarks prove that a AthlonXP 1800+ has the same performance as a P4 1.8Ghz.

I've read them, have you?

 

[RedShirt]

<< Seems you have not got the intelligence to read Anandtech's own reviews and benchmarks for yourself, as i've said counless times the PR system is not biased as even the benchmarks prove that a AthlonXP 1800+ has the same performance as a P4 1.8Ghz.

I've read them, have you? >>



OK, the whole entire point of this thread has been this:

1. Intel has released newer processors that run with a higher FSB
2. If AMD keeps the PR rating system as they do now, a PIV 2400 will beat an Athlon 2400+ in almost everything
3. This would be very deceptive on the part of AMD since the processor is not really 2400 MHz and they would be claiming it is 2400+

Even if they change the scale, this would be bad! For instance, at least when using a REAL MHz rating, you know a Pentium IV 2400 will outpreform a 2000, the bigger the number in MHz in the same family of processors, means more preformance (unless the bus speeds are different, but Intel usually denotes this with a letter in the processor's name).

Enter AMD, who know has this crazy scale they invented, let's say the change the scale to accomodate these newer Pentium IV's, let's say they need to have a processor that is 66 Mhz faster than in the normal scale to compete... So, somewhere along the line, they are going to have a XP 2200+ or so that is on the old scale and then a 2300+ (these are just examples) that is on the new scale. There will be a difference of over 133 MHz in reality in these chips, when before, the difference had only been 66 MHz.

You may not see the problem with this, but when you are using PR ratings, this kind of thing is VERY deceptive, one would not suspect there to be a 133 mhz difference between a 2200+ and a 2300+ especially when there is only a theretical differance of 100 MHz, one would believe, since AMD chips are faster, clock for clock, that the difference would be less than this.

This leads to other questions, such as, when Intel changes their chips again, will AMD simply continously change their scale? That is utter chaos and the need to totally ditch the scale or have an independant company come in and give PR ratings to every processor exists.

The arguement of this thread has never been about how the PIV uses its bandwidth and how Quake3 likes the Pentium IV, it's about the flaw in the PR rating that AMD just decided to make up one day.

 

[grant2]

To say that a RISC CPU is more efficient because it processes less instructions per cycle seems to go against the
very logic of using RISC, as far as I understand it.

That's why mac zealots stopped beating their chests a few years ago about how "x86 is at the end of its life" and "intel will have to switch to risc".... but now they're quiet since x86 architecture is *still* wiping the floor on their precious "risc" architecture and is in fact ADDING instructions (3dNow!, SSE, etc.)

Risc just doesn't make sense in a general-purpose (PC) system.

 

[LikeLinus]

<<Thats a typo stupid, you should have pick that up yourself>>
Hence the LOL, it was sarcasm and pretty funny.

<<See my link above, games that are CPU and FPU intensive the Athlon wastes the P4, plan and simple>>
Wait, what does FPU have to do with the bandwidth of RDRAM? You are mixing two different performance measures. No one said that clock for clock the AMD can beat Intel. Dude you can't even keep your arguments straight.


<<Ask yourself why in bechmarks with Quake3 engine(in Anandtech own reviews) the P4/RDRAM is able to perform better than AthlonXP nearly all the time(forgeting that P4 is 100's of Mhz ahead)>>

LOL this kid is a trip. He thinks the best benchmarks there are - are games.

FYI Quake 3 has nothing to do with bandwidth...ugh. It has to do with it's optimization of SSE and Intel chips in general.


<< It's been proved that the P4's quad pumped bus and it's huge bandwith is pritty useless >>

See same reviews on P4.

PLEASE show me in that article where it says that Intels Quad Pumped 100mhz FSB is useless and prove (as you've stated).

<< AMD know that huge bandwith figures mean nothing and in reality >>

AMD's own FSB(266Mhz DDR) proves you dont need heavy bandwith for todays and tomorrows games, apps to perform on par with a 400Mhz QDR FSB of the P4 and RDRAM. >>

Man, this is starting to get silly. AMD's performance in games has NOTHING To do with it's DDR "owning" RDRAM. It's all about FPU's and graphics cards. As Stated, it is industry known that RDRAM is a superior ram standard, but poorly implimented because of RAMBUS. If bandwidth didnt matter so much, why does AMD support DDR? You don't need DDR to run over 100 FPS on Quake 3. Your argument has so many holes.

That and the fact you base all your "knowledge" and "proof" off one Anandtech article from November 20th, 2000 , LOL. Dude get with it.


If you really go READ THE ARTICLE....it even states why the AMD is a better performer than the P4

"Remember that the Athlon can do more in a single clock than the Pentium 4, making the 300MHz difference in clock speed between the Pentium 4 and the Athlon compared here mean very little." THIS has NOTHING to do with Bandwidth

Lets check this article out.

http://www.anandtech.com/cpu/showdoc.html?i=1615&p=2 This is about the new 133mhz quad FSB for the 2.4 (and 2.53 chips). Read through it and see if you can see the benifits of using a higher FSB.

 

[TSDible]

LikeLinus is completely correct.

The AMD PR system is biased. If you say it is not, then consider this...

Would AMD have EVER created a PR system in which a 1800+ cpu did not perform at least as well or better than an Intel 1.8GHz CPU. The answer is clearly no. To do so would do more harm than good from a business point of view.

AMD creating their own PR system is practically the definition of bias...

 

[BlvdKing]

You guys are all dorks.

 

[formulav8]

<< <<The AthlonXP 1800+ in benchmmarks is able to out perform the P4 1.8Ghz, HOW SIMPLE DO YOU WANT IT?

AMD only went from 1.43Ghz(1600+) to 1.47Ghz(1700+) because that jump was equivalent to a T-Bird 100Mhz jump. >>

LOL That has absolutely NOTHING To do with the stuff you were talking about. You are stating that a p4 using RDRAM loses to an AMD XP using SDRAM.

PROVE THESE BS STATEMENTS YOU ARE MAKING. LISTED BELOW

1. <<RDRAM can but against your standard AthlonXP/SDRAM it's not very good at all.>>
2. <<Fact is not many apps or game take advantage of RDRAM's huge bandwith> LOL What a joke.
3. << It's been proved that the P4's quad pumped bus and it's huge bandwith is pritty useless>>
4. <<AMD know that huge bandwith figures mean nothing and in reality>>

Dude I hope you don't believe the BS spewing from your mouth. I also hope to your parents are wearing boots right now.

How else are AMD supposed to tell the customer? Well ever heard of advertising and marketing? Commericals? Ads in magainzes? Having PCWorld do testing? They don't have to create some magical rating scale to lie to customers and misguide them in buying a product. No one is saying it isnt as fast. Still you miss the point.

AMD should be trying to advocate for a real PR system. THESE are the things they should be doing.

<Athlon 1.2Ghz/RDRAM> When did AMD start using RDRAM? lol. Also you do realize that the P4 losing to the AMD has really nothing to do with the ram it uses. Its the pipelines that cause the P4 to slow down. Intel lengthened them so they could scale higher becaues AMD started a mhz war. Now AMD can't keep up, and they wanna play this whole "oh mhz dont matter". Ignore the fact that AMD was the one with commercials saying that they were the first with a 1ghz chip and all the ad's. lol

You also realize that most industry people consider RDRAM superior to DDR? But you can still sit around and tell everyone that it's "useless". hehe

I nominate you for noob of the year. >>




What is with your name calling. You act like Intel is your mother or something. Its only his opinion. He's not threatening your life.

 

[imgod2u]

<< AMD did only have 2% market share, but it rose drastically when the highly clocked K6 was released. Intel purposely engineered the Pentium 4 for high clock speeds, and in my opinion, they did it for marketing reasons. If you don't agree, that's your opinion and you're entitled to it, but don't call mine BS, or else I will argue with you untill I die.

P.S. I am not and AMD fanatic. I buy what is best at the time, performance and budget wise. That is why I just ordered a P4B533, P4 1.6A, and PC3000 DDR. AMD has dropped the ball, and will not regain it untill Hammer. >>



I would think your preference in companies or processors shouldn't really affect your conclusion on why Intel decided to go for the P7 core design. As I've stated in another post, there are other advantages asside from just the huge jump in clockspeed with a 20-stage pipeline. I'm sure that the increase in the clockspeed was a reason for the early release and the early dump of the P6-based chips (the P3 had some headroom, Intel could have just put in copper interconnects and do put a DDR bus it, it still could've competed with the Athlon). However, the original design? Why would a company controlling 97% of the market, with the competition not gaining any marketing shares (with no money to advertise) even care about what its competition offers? I'm not arguing the guys in marketing didn't consider a MHz monster, but this implication that that was the only reason is very rediculous as far as I'm concerned.

As far as the FSB/memory bandwidth, as I've pointed out in another post, the required memory bandwidth depends on MHz alone in current processor designs. A 2.4 GHz processor will require 2.4 billion instructions to be fetched every second, no matter what the average IPC is. The P4 will inherently need more memory bandwidth and can potentially do more with it (not saying that it actually has). One of the reasons for the P4's low IPC is that the memory bandwidth can't fetch enough instructions per second to keep the pipeline full. If you put a very fast FSB and memory on the Athlon, it will not benefit as much as a P4 would from the added memory/FSB. In other words, memory and FSB performance is not something that will directly affect overall performance separate and independent of the processor. So just because you increase a processor's FSB/memory doesn't mean its performance will increase. So if the current FSB/memory on the Athlon supplies 1.73 billion instructions per second, then adding an FSB and memory as fast as that on the P4 will not help. On the P4 however, seeing how the increase to a 133MHz FSB and 533MHz memory helps its performance significantly, I'd say its current FSB/memory (100/400) doesn't provide it with enough instructions per second and hence, does hinder it.

 

[absolutiza]

yea.. i agree

 

[MadRat]

<<One of the reasons for the P4's low IPC is that the memory bandwidth can't fetch enough instructions per second to keep the pipeline full. If you put a very fast FSB and memory on the Athlon, it will not benefit as much as a P4 would from the added memory/FSB. In other words, memory and FSB performance is not something that will directly affect overall performance separate and independent of the processor.>>

Thats not true. The overclockers pushing towards 200fsb are getting better benchmarks then overclockers pushing at a standard 133fsb and simply raising the multiplier. The Athlon has probably been cache starved alot longer than people realize. The 64-bit pathway to the core from the L2 cache is its worst limiting factor as speed increased above 1GHz. AMD should have probably instituted a 128-bit pathway or better on the newer designs. As it is the Athlon has changed very little from the original slot-1 version to the current Palomino.

 

[wfbberzerker]

i tend not to agree with stuff posted on toms hardware (in my experience, the site has always been intel biased)

 

[christoph83]

Thats not true. The overclockers pushing towards 200fsb are getting better benchmarks then overclockers pushing at a standard 133fsb and simply raising the multiplier. The Athlon has probably been cache starved alot longer than people realize. The 64-bit pathway to the core from the L2 cache is its worst limiting factor as speed increased above 1GHz. AMD should have probably instituted a 128-bit pathway or better on the newer designs. As it is the Athlon has changed very little from the original slot-1 version to the current Palomino.

I think this was his point.....

If you put a very fast FSB and memory on the Athlon, it will not benefit as much as a P4 would from the added memory/FSB.

If you were running a 2ghz northwood at 10x200 with RDRAM, the P4 would benefit tremendously, whereas the athlon at 200FSB would not get as large of a boost, but would still get a boost.

 

[imgod2u]

Originally posted by: MadRat
<<One of the reasons for the P4's low IPC is that the memory bandwidth can't fetch enough instructions per second to keep the pipeline full. If you put a very fast FSB and memory on the Athlon, it will not benefit as much as a P4 would from the added memory/FSB. In other words, memory and FSB performance is not something that will directly affect overall performance separate and independent of the processor.>>

Thats not true. The overclockers pushing towards 200fsb are getting better benchmarks then overclockers pushing at a standard 133fsb and simply raising the multiplier. The Athlon has probably been cache starved alot longer than people realize. The 64-bit pathway to the core from the L2 cache is its worst limiting factor as speed increased above 1GHz. AMD should have probably instituted a 128-bit pathway or better on the newer designs. As it is the Athlon has changed very little from the original slot-1 version to the current Palomino.

I'm not saying there wouldn't be a performance increase, I'm just saying not as much. There are still, of course, instances in which memory is the limitation (it does other things asside from fetch instructions) but for the majority of cases, the Athlon is capable of fetching enough instructions every second (1.73 billion sounds about right for the fastest Athlon) while the P4 is still straining for more fetched instructions per second (just look at how much a faster FSB and memory helped it).

 

[Degenerate]

In short. Pr was invenented for marketing purposes. It creates a good image of the Athlons, while being confusing.

 

[MadRat]

<<I'm not saying there wouldn't be a performance increase, I'm just saying not as much. There are still, of course, instances in which memory is the limitation (it does other things asside from fetch instructions) but for the majority of cases, the Athlon is capable of fetching enough instructions every second (1.73 billion sounds about right for the fastest Athlon) while the P4 is still straining for more fetched instructions per second (just look at how much a faster FSB and memory helped it).>>

The problem is that the Athlon cannot fetch enough instructions per second to keep it busy. If there is a performance increase from raising the FSB (but maintaining the same overall MHz) then it proves my point, because the only part of the processor running faster is the 64-bit pathway to the processor from the external memory. The Pentium 4 scales better in performance as the FSB is raised simply because it is MORE cache starved.

 

[imgod2u]

The problem is that the Athlon cannot fetch enough instructions per second to keep it busy. If there is a performance increase from raising the FSB (but maintaining the same overall MHz) then it proves my point, because the only part of the processor running faster is the 64-bit pathway to the processor from the external memory. The Pentium 4 scales better in performance as the FSB is raised simply because it is MORE cache starved.

As I mentioned, there are other things that require memory/cache transactions asside from instruction fetching. I'm sure there may be some cases in which not enough instructions can be fetched, but for the most part, I think the Athlon's FSB and cache supplies it with somewhat decent throughput to fill its pipeline.

 

[jhu]

As stated serveral times. There needs to be an independant, non-bias, board of members put together who can create a set of benchmarks and scale for the performance of chips. This is just like the age old question of whos faster....Apple chips or PC chips. If there was a standard, all this would be solved.

yeah, it's called spec. but who wants to wade through a bunch of spec benchmarks just to see what cpu is better (which,btw, is still the power4)?

 

[mechBgon]

It would be impossible to make a meaningful universal performance metric for CPUs with such diverse capabilities. Everyone in this thread knows that a CPU can perform well at one thing and poorly at another. To take a very simple example, look at the original Pentium4's. If you judged by their Quake3 performance (using RDRAM, the only option at that time), they were the best CPU. If you judged by their RC5-64 performance, they were an utter joke, easily beaten by a Celeron II running at half the MHz. Whether it's fast or slow depends upon what you're doing with it.

Another example would be rendering in 3D modelling software. 3DSMax favors the strong FPU of an AthlonXP, Lightwave 7 favors the SSE2 capabilities of a Pentium4. A universal performance mark would be a disservice in that instance.

Of course, the person building a system for 3DSMax or Lightwave 7 will do enough research that a PR rating is a moot point for them. The PR system is, as has been said, aimed at the average consumer, who doesn't know that not all MHz are created equal.

As far as the average consumer goes: for a while, I held the opinion that AMD should play their underdog role to the fullest, by rating their CPU at its true MHz and emphasizing their lower prices.

However, with the release of the "Celeron4," armed with only 128k of L2 cache and probably paired with low-performance platforms that use some of the memory bandwidth for onboard video, I think it would be a joke to compare a 1.8GHz AthlonXP with a 1.8GHz "Celeron4" using SDRAM or even DDR. Let the PR system stand, and realize it works both ways... for AMD's PR system to be accurate toward a Celeron4, they would have to jack their ratings by several levels! See what I'm saying?

And RedShirt, as for your claim that the PR system doesn't hold true for the Pentium4-B family, I have a one-word answer: "Clawhammer." Consider it a recalibration-point for the PR system.

 

[AGodspeed]

Well said mechBgon.