But how does it affect you? Answer: it doesnt. You look at the benchmarks, you decide. It's not a low method, it's marketing. So what has AMD done that requires forgiveness?
I see, so a 2000+ Palomino runs at 1667MHz a 2100+ Palomino runs at 1733MHz and a 2200+ Thoroughbred runs at 1800MHz. The added "couple of hundred points" are... where, exactly?
So A Willamette 1.6 being trounced by a PIII 1Ghz and indeed a Willamette 1.8 by a Palomino 1.2 (Ghz) was consistent? The PR system is far from perfect, but at least you can safely say that an XP2000 is roughly equivalent to a P4 2GHz. It's not entirely consistent but it gives AMD a chance in the retail market.
So would you take a Wilamette 2GHz over a Palomino 1.6GHz? The clock speed is 25% higher so it must be better
Hey guys, if you all think IPC is so much more important than clock speed and the PR system is completely legit, then you know that the offical line from AMD is that the PR rating indicates how fast a Thunderbird would have to be to perform at the same level as the same Palomino. So a PR 2000+ means the Palamino 2000+ performs the same as a Thunderbird 2GHz. Thats funny, because we all know that in reality the 2000+ performs about on par with a P4 2GHz. Obviously that means that the P4 and the Thunderbird have the same IPC. Basic logic here people, logic that backfires in AMD's face, and proves the PR system to be completely fraudulent, since we know the Tbird has a higher ICP than the P4.
Kramer
Kramer
Nope, important when comparing two CPUs of identical clock speed
Said by...?
Very true. AMD cannot claim that it is the same as a P4 2GHz because it would result in legal action. And with XP vs P4, some applications favour one or the other, so it's hard to draw a direct comparison.
Roughly so, in certain applications, under certain conditions.
So the Palomino PR is "On a Par" with the T-Bird MHz and is also "Rougly the same" as the P4 MHz and we can draw definitive conclusions from this? It's marketing. It's advertising. It's not an exact science. To call it fraudulent on the basis of two different imprecise comparisons is pointless. Why do you even care - IT DOESN'T AFFECT YOU!
If its completely imprecise then how is it not fraudulent? Anyway, It's not a matter of directly affecting anyone, espicially anyone on this forum, but it is a matter of principle. You seem to think that it's "unfair" to AMD that Intel's chips scale higher in MHz, and I can't understand that sentiment. If they feel like they need to show people that their chip is simply engineered differently, then they should at least have used an industry standard, industry approved benchmark or naming scheme. I don't have a problem with the idea of a PR system if thats the way it's gonna be, but AMD's system is biased, misleading, and barely holding on to any connection to facts.
All I know is, if i was Joe Consumer, and bought a 3400+ system only later to find out it was running at 2GHz, I would be pretty upset and confused, no matter how well the system performed.
Kramer
Adul
Elite Member
one wonders how intel will market the banyas cpu. It is a lower clockspeed then the P4, but has a much high IPC. and lower power usage.
Yea, most consumers are stupid. But PR rating, IMO, is messed up. Its kind of like wattage ratings on stereo systems.
Right now there are a lot more factors than Mhz for cpu performance. Frankly I think my 2.4 Ghz Pentium4 northwood with PC1200 RDram at 600Mhz FSB should be a pentium4 2700+
because obviously it can beat a 2.4/400 Ghz p4 running PC2100 DDR.
Now if you factor in SSE2, a Pentium4 1.6 Ghz can BEAT an AthlonXP 1.6Ghz (2000+), clock for clock. So much for that PR system. But in some OTHER applications, an AthlonXP 1.6Ghz can beat a 3Ghz Pentium4. There is way too many variables to determine an accurate PR rating, but so far that PR rating is proving to be pretty accurate.
Right now there are a lot more factors than Mhz for cpu performance. Frankly I think my 2.4 Ghz Pentium4 northwood with PC1200 RDram at 600Mhz FSB should be a pentium4 2700+
because obviously it can beat a 2.4/400 Ghz p4 running PC2100 DDR.Now if you factor in SSE2, a Pentium4 1.6 Ghz can BEAT an AthlonXP 1.6Ghz (2000+), clock for clock. So much for that PR system. But in some OTHER applications, an AthlonXP 1.6Ghz can beat a 3Ghz Pentium4. There is way too many variables to determine an accurate PR rating, but so far that PR rating is proving to be pretty accurate.
It's all very well saying we need an industry-wide standard for performance but as dexvx says, different applications favour different CPUs so an entirely objective approach is almost impossible. If you compare the clock speeds of the palominos with the P4s, the XP performs far better at a given clock speed. Comparing the PR rating with the P4s clock-speed gives a much better indication of relative performance. It's not as if there's no parity there. AMD could happily have made the 2200+ into the 2500+ to compete with intels high end processors but they didnt.
No system is perfect, but the PR ratings give the general public a far better comparison than the clock speed would.
But the man on the street doesnt care about the contents, he cares about the speed. If he buys a 3400+ system that runs at about the same speed as the man-next-door's 3.4GHz P4 he's perfectly happy.
Ah, but everyone knows that the stereo system wattage is meaningless.
I had someone trying to tell me that his "400W stereo multimedia speakers" are better than my 100w KEF standmounters. there are two completely different systems there, RMS and PMPO just to confuse things.And that's why any complete objectivity is impossible. I think the PR ratings give AMD a chance of being decent competition for Intel in the retail market and that has to be a good thing.
I own an athlon system right now, but my next system will definitely be INTEL! I just find the rating scheme that AMD uses a step back into the K6 days. I mean comn, give me the correct Mhz rating instead of 3400+ or some bs.
I think that the PR rating might end up boxing AMD in a corner with no way to get out.
That said, to Joe Consumer what looks faster on paper a AMD Athlon 1800Mhz or a Pentium 2200Mhz? Your average store bought comsumer has been conditioned to think that mhz is king. Hence the PR rating.
Now most of us know that so far AMD has been a little bit conservative with their PR system, as long as they keep that up I will have no problem with it. Even though I think it would be a little disheartening to have to name all of my products according to how the competition is performing.
Will they always be 1 step behind Intel to make sure they have the right PR rating in place?
3400+ seems a little ambitious to me, they might step in it this time. Hopefully not though because competition is a good thing.
That said, to Joe Consumer what looks faster on paper a AMD Athlon 1800Mhz or a Pentium 2200Mhz? Your average store bought comsumer has been conditioned to think that mhz is king. Hence the PR rating.
Now most of us know that so far AMD has been a little bit conservative with their PR system, as long as they keep that up I will have no problem with it. Even though I think it would be a little disheartening to have to name all of my products according to how the competition is performing.
Will they always be 1 step behind Intel to make sure they have the right PR rating in place?
3400+ seems a little ambitious to me, they might step in it this time. Hopefully not though because competition is a good thing.
AMD did not create the PR system to decieve people. They created the PR system out of neccesity. Before the PR system came out, AMD had to sell it's 1.4Ghz Athlon for less than the price of a 1.5Ghz P4. AMD's ASP's were pushed down drastically because consumers buy products based on Mhz, so they had to price their parts according to Intel. Without the PR system, AMD would have gone under, because their ASP's would have been so ravaged. Whether you like it or not, the PR system is here to stay. Without AMD you would be paying a lot more for the P4. Too bad so sad, get over it.
When I mentioned ?what the PR rating may become?, that?s exactly what my concern is. Are they going to start ?stretching? the numbers, oh because Intel is getting better yields?!
So when Intel comes out with a 4.0GHz part, and AMD can only manage a 3.0GHz unit, are they going to call it 4000+?
Again, I?ll have to see a review of an alleged AMD 2.6GHz being able to perform like a true 3.4GHz before I believe it. Then again if they can produce a 2.6GHz, then they should market it like a 2.6GHz, I?m sure the price itself will catch peoples attention.
I think it?s unfortunate AMD can't get better yields, they seem to be doing a good job of keeping up when it comes to product design.
Hehehe? and in conclusion it?s obvious some agree with AMD?s PR scheme and some don?t.
EDIT: OOPS...
I'm a bit curious where you got the impression that a 2.6GHz Hammer will cost anything less that $500. All indications point to the new Hammers being priced similarly to the highend P4s.
Kramer
I'm sorry but their is no maybe about it! Hammer is simply a better design than P4 and will rule the performance sector. I think that AMD can capture the performance crown with the Athlon XP 2400+ if they released it in sept. !
Personally, I think the PR+ rating is incredibly lame and stinks of Cyrix everytime I see it.
That said, I can understand that AMD MUST use it if they have ANY chance of competing with Intel.
You know and I know that Maw and Paw Smith buying a computer for little Johnny doesn't know or care jacksh!t about IPC.
These aren't necessarily *dumb* consumers. They're simply non-technical.
For example, I know jacksh!t about cars. To me horsepower is king. The higher the number the faster it goes as far I am concerned... when in reality there are a LOT of factors in play.
AMD's PR rating, while icky, is a neccesary business decision and doesn't in any way affect my opinion of thier processors.
That said, I can understand that AMD MUST use it if they have ANY chance of competing with Intel.
You know and I know that Maw and Paw Smith buying a computer for little Johnny doesn't know or care jacksh!t about IPC.
These aren't necessarily *dumb* consumers. They're simply non-technical.
For example, I know jacksh!t about cars. To me horsepower is king. The higher the number the faster it goes as far I am concerned... when in reality there are a LOT of factors in play.
AMD's PR rating, while icky, is a neccesary business decision and doesn't in any way affect my opinion of thier processors.
so many people here that are saying that the pr system sux seem to either forget or not know that AMD never said that the pr system was a comparison to the P4 chips.....that's now what it is derived from.........just b/c amd and intel are the main competitors doesn't mean that amd's comparison system is comparing to the P4.......how is it misleading when you can't go buy a 2ghz thunderbird?.....u can however buy an xp 2000+ which will be equivalent to a 2ghz thunderbird if such an animal existed.....the PR system has nothing to do with intel.....it's not amd's fault that the average consumer looks at it as a AMD vs. Intel comparison.........they clearly stated what it was for as well as why they did it
Thats all fine and good, but the Palamino is just a tweaking of the Thunderbird design, so according to your logic, Intel could say the P4 2.53B is a "2800+" because it would take a 2800MHz Willamette to equal the performance of a 2.53GHz Northwood. Jeeze, if that ever happend i'd hate to see the outrage in here, even thought it's
exactly the same thing that AMD is doing.
Kramer
While that's true, and a valid point I don't think it makes much difference what the Mode rating is supposed to be compared to. The consumer will compare it to other products on the market, the only other products at comparable MHz/PR Rating are Intel's Willamette/Northwood Pentium 4 processors, and Intel's Willamette-128 Celeron processor.
If the Model rating doesnt match up to those processors clockspeeds then AMD will quickly get an awfully bad reputation regardless of whether the Model rating is based off the Pentium4/Thunderbird/i80486 or anything else.
As for my opinion of the Model rating- I view it as a necessary evil. Yeah it stinks, and it's far from my ideal measurement of performance.
But at this point it's an absolute necessity.
AMD, Cyrix, IDT etc... in the past they've all tried to teach consumers that MHz does not equate to performance, all attempts have failed. AMD cannot afford to market the AXP 2200+ by it's clockspeed- 1.8GHz. If they were forced to market processors by clockspeed their marketshare and ASP's would drop extremely rapidly. Their mainstream AthlonXP 1800/1900+ processors would be laughed off the market if they were forced to compete on the merits of clockspeed alone.
In the end, I view the Model rating as an unfortunate necessity.... so long as the Model rating remains realistic compared to Intel's highest performance X86 processor at a comparative MHz/Model I will accept it.
Right now, I do view it as completely realistic. If the day comes that AMD markets a Model rating that is NOT on par with Intel's competing solution then I will be among the first to speak against AMD.
Think of it this way: what if Intel and AMD went in opposite directions in which the AMD chip is the mega pipelined high frequency CPU and the Intel chip is this more efficient design that squeezes more performance per clock cycle. Think of the market position AMD would enjoy, not only are our chips faster, they cost less too!!! I would be willing to bet Intel would come out with some sort of marketing term for the performance characteristics of their chip.
In a way I hope PR ratings do succeed because if it doesn't, whats to prevent AMD from redesigning their chips where the main focus is frequency and not true performance. Its perception, but I would rather have a cpu with real performance then perceived performance.
In a way I hope PR ratings do succeed because if it doesn't, whats to prevent AMD from redesigning their chips where the main focus is frequency and not true performance. Its perception, but I would rather have a cpu with real performance then perceived performance.
In case people haven't noticed, Intel's method of "cheating" for market share (designing a CPU that is able to scale up quickly in clockspeed) is not cheating and it's not a marketing gimmik: the chips are FAST. There is no issue of "percieved performance" anymore, because obviously Intel chose a design that is more power and less hype, despite what many people believe. Sure at first it was disappointing, but now the true purpose of the design is just starting to show, and anyone that thinks all Intel wants is to sell MHz is clearly mistaken. Right now, Intel is making the fastest chips in the world, both in MHz and in pure power. We'll see what AMD can do with the Hammer. Hopefully we'll get some real numbers soon. Things can only get faster, afterall.
Kramer
So speaks Rand
The point at the moment is that the ratings <are> fair. They give a good indication of the equivalent P4 performance to a palomino part.
The point is that intel are 750+ MHz above AMDs highest-clocked part. The PR ratings are to give the Palominos parity with the P4. Sure intel could name their parts the 2800+ or somesuch, but it'd be pointless since they're already ahead in the MHz stakes. It's a marketing gimmick to give AMD a chance to sell equivalent performance at a lower clockspeed.
It'd be interesting to see how 64bit will play into this. In reality, a 64bit 3000+ (if the PR system is on par with the XP's PR system) wont any different than a 32bit p4 3.06 Ghz. Now i wonder if consumers think 64bit 3000+, it'll be twice as fast as a 32bit 3.06Ghz.
A lot of people even here think 64bit is the bomb also. But 64bit is actually slightly slower than a 32bit counterpart, if they ONLY do 32bit operations. Now you'll probably see Intel sporting Pentium4 3.06Ghz 64bit+, because technically with SSE2 extensions, the pentium4 can also process in 64bit. Although it isnt truly a 64bit chip, it can still do 64bit operations, just like even though the XP isnt truly a 2Ghz chip, it is up to par with 2Ghz chips.
A lot of people even here think 64bit is the bomb also. But 64bit is actually slightly slower than a 32bit counterpart, if they ONLY do 32bit operations. Now you'll probably see Intel sporting Pentium4 3.06Ghz 64bit+, because technically with SSE2 extensions, the pentium4 can also process in 64bit. Although it isnt truly a 64bit chip, it can still do 64bit operations, just like even though the XP isnt truly a 2Ghz chip, it is up to par with 2Ghz chips.
I am in the process of doing the same thing. This is my first and last homebuilt AMD system. Before I got into building computers I bought a k6-2 350 that was a big POS. One of the reasons I'm going back to intel is they run coooler. Another reason is stability. Once I get the P4 built(all I will need after monday is the CPU) this will be used for storing files.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter