AMD’s Executives Admitted Inability to Compete – Intel.

[nyker96]

I don't blame him for this. It is true at least until BD shows up next year.

http://www.xbitlabs.com/news/cpu/di...ives_Admitted_Inability_to_Compete_Intel.html

 

[EarthwormJim]

Processor speed isn't the main issue, it's having a complete platform.

AMD seems to be remedying this which may also have motivated their Ati purchase so they could have a complete chipset/graphics solution.

 

[Jeeebus]

http://forums.anandtech.com/showthread.php?t=2041065

 

[frostedflakes]

In 2004 it wasn't so true, which is when the guy said this. Kind of makes you wonder if he knew what the hell he was talking about, because back in 2004, AMD was near its peak. K8 chips in desktops and servers were outperforming almost everything Intel had to offer, and were doing so while using less power. I mean if he had said it in 2002 I could understand, but anybody who would say something like that during 2004 must have been woefully uninformed. 2004-2006 were the good years for AMD.

Also, people shouldn't get their hopes up for BD IMO. I'd love to see it be a major success for AMD, but it remains to be seen how competitive it will be with Intel processors. Also, it's not like Intel is going to be sitting on their hands for the next year, they're going to be improving their processors as well. Just don't want people to be disappointed if it ends up being a flop. It does sound promising, though.

 

[dguy6789]

The Athlon 64 and Opteron launched in 2003. The Core 2 launched in 2006. From the time of the Athlon 64 launch until the time of the Core 2 launch, AMD had: The performance crown, the lowest power consumption at every performance level, and the fastest processor for every given price segment. This was true for both the desktop and server market.

 

[zsdersw]

From the time of the Athlon 64 launch until the time of the Core 2 launch, AMD had: The performance crown, the lowest power consumption at every performance level, and the fastest processor for every given price segment. This was true for both the desktop and server market.

Wrong. The Northwood version of Netburst was very competitive with the A64 in most areas and was faster than it in a few. It was also pretty miserly on power consumption.

 

[dguy6789]

Wrong. The Northwood version of Netburst was very competitive with the A64 in most areas and was faster than it in a few. It was also pretty miserly on power consumption.

It was not competitive. The fastest Northwood was "sometimes" as fast as an Athlon 64 but that was not commonplace. The price of said Pentium 4 was always a LOT higher than the price of the Athlon 64 it just barely managed to nudge out. For it to be competitive, it would have to be similar performance(which it was not, it was slower way more often than not) and similar price(it was way more expensive). It would have been very unwise to buy an Intel setup when an Athlon 64 was available. Northwood was also discontinued at the very beginning of 2004 and that was the closest Intel ever came to AMD for that generation.

"But make no mistake, what you see before you is not the power hungry, poor performing, non-competitive garbage (sorry guys, it's the truth) that Intel has been shoving down our throats for the greater part of the past 5 years." -Anand talking about how Conroe differs from the Pentium 4

 

[zsdersw]

It was not competitive. The fastest Northwood was "sometimes" as fast as an Athlon 64 but that was not commonplace. The price of said Pentium 4 was always a LOT higher than the price of the Athlon 64 it just barely managed to nudge out. For it to be competitive, it would have to be similar performance(which it was not, it was slower way more often than not) and similar price(it was way more expensive). Northwood was also discontinued at the very beginning of 2004 and that was the closest Intel ever came to AMD for that generation.

Sure it was competitive. Media encoding almost always ran faster on a P4 than an A64.. and not by a little amount, either.

Your overall statement is wrong in its explicity. You use absolute terms where absolute terms do not belong.

 

[mav451]

I think more than anything else it was the volatility of chipset/sockets. You had 940, which required registered, 754 which was EOLed almost before it began, and then 939 which ended up being EOLed too quickly. I mean, great that it had a performance crown, but that was just horrible execution and probably contributes to the validation issue.

I remember b/c I helped people build with the NF3 250 chipset, but ended up going NF4 Ultra on the s939 just a bit later. If it was tough for me, I can't imagine for less technically gifted people to figure out what the hell was going on.

 

[Maximilian]

Sure it was competitive. Media encoding almost always ran faster on a P4 than an A64.. and not by a little amount, either.

Your overall statement is wrong in its explicity. You use absolute terms where absolute terms have no basis in reality.

Yeah but judging a processors overall competitiveness on media encoding alone back then was stupid and remains so today lol.

P4 traded some blows with A64 but overall it was outclassed in almost every area. Dont get me wrong back in 2004 a P4 3.2C was a very good cpu to have, its just somone with an A64 3200+ had a better cpu.

 

[dguy6789]

Sure it was competitive. Media encoding almost always ran faster on a P4 than an A64.. and not by a little amount, either.

http://www.anandtech.com/showdoc.aspx?i=2668&p=7

I see the Pentium winning in 2 of 4 media encoding tests compared with the Athlon 64. There's also no way you can call the amount the P4 wins by anything other than a little amount.

Your overall statement is wrong in its explicity. You use absolute terms where absolute terms have no basis in reality.

Absolute terms work fine when it's the truth. From 2003-2006, there was never a reason to buy a Pentium 4 and that's the truth.

 

[Hacp]

Sure it was competitive. Media encoding almost always ran faster on a P4 than an A64.. and not by a little amount, either.

Your overall statement is wrong in its explicity. You use absolute terms where absolute terms do not belong.

2005 dual core came out. Most of 2005 and 2006 was dominated by AMD.

 

[Nemesis 1]

I can't believe this was posted here at AT . Didn't the OP read the XS topic .If ya want something good to talk about Intel posted earnimgs yesterday of Over 2 billion after they paid AMD 1.25 Billion Record earnings for Intel Now thats interesting

 

[zsdersw]

http://www.anandtech.com/showdoc.aspx?i=2668&p=7

I see the Pentium winning in 2 of 4 media encoding tests compared with the Athlon 64. There's also no way you can call the amount the P4 wins by anything other than a little amount.

Half of the tests is a little amount? Your referenced Anandtech article also shows comparisons with only Smithfield/Presler CPUs, not Northwood.

Absolute terms work fine when it's the truth. From 2003-2006, there was never a reason to buy a Pentium 4 and that's the truth.

Yeah, but it's not the truth.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=1956&p=25

When you include AMD in the picture, the recommendation hasn’t changed since the Athlon 64 was introduced. If you find yourself using Microsoft Office for most of your tasks and if you’re a gamer the decision is clear: the Athlon 64 is for you. The Pentium 4 continues to hold advantages in content creation applications, 3D rendering and media encoding; if we just described how you use your computer then the Pentium 4 is for you

 

[zsdersw]

Yeah but judging a processors overall competitiveness on media encoding alone back then was stupid and remains so today lol.

I wasn't judging anything, merely stating how wrong it was for him to use such explicit and absolute terminology. He said there was no reason to buy an Intel CPU during the Netburst era. Clearly that's wrong.

 

[dguy6789]

Half of the tests is a little amount?



Yeah, but it's not the truth.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=1956&p=25

And at what cost? Performance without price is a worthless comparison. You've consistently ignored price since the beginning of this discussion. If the Pentium 4 was 5 times faster than the Athlon 64 but was a million dollars, it would still be a piece of trash. You can't ignore price. Sure, a 3.2Ghz Pentium 4 might beat an Athlon 64 3200+ in a couple tests, but you know what? That doesn't matter at all and doesn't in any way disagree with anything I've ever said. Do you know why? A 3.2Ghz Pentium 4 was more than one hundred dollars more than an Athlon 64 3200+. A 3.8Ghz Pentium 4 would be faster than an Athlon 64 3200+ also. Why aren't you making that comparison? Because it's a lot more expensive. But so is the 3.2. Compare the Athlon 64 to something from the same price.

When a 2.8Ghz Pentium 4 consistently matches or beats an Athlon 64 3200+ and a 3.4Ghz Pentium 4 consistently matches or beats an Athlon 64 4000+ then you can claim the Pentium 4 was competitive because those processors were the same price. The difference increases when you move to dual core. A Pentium D 840EE(2x3.2Ghz) was MORE expensive than the Athlon 64 X2 4800+(which was two Athlon 64 4000+ processors). At every single price point, the Athlon 64 was unrivaled, which is what I said at the beginning and has always been true.

The only time you find the Pentium 4 winning is if you compare based on performance only which should never be done.

When Anand says something like the Pentium 4 is better at a certain task, he is saying it as if one is comparing a 3.2 P4 to a 3200+ 64.

 

[Arkaign]

dguy, I think in your zeal you've forgotten that AMD charged dramatically higher prices, particularly during the X2 era, when they firmly held the performance crown. The only bargains from AMD to be had were the Semprons and lower-tier Socket 754 Athlon 64s.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2410&p=1

4800+ $1001
4600+ $803
4400+ $581
4200+ $537

??? Value not found.

http://www.legitreviews.com/article.php?aid=226

"With AMD's entry level dual core processor costing more than double that of Intel's, many consumers simply can't afford AMD's Athlon 64 X2 processors."

Even the eventual launch of the 3800+ X2 at $354 wasn't all that affordable for most people. A lot of folks like me picked up things like the 805D and clocked them up in the 3ghz+ range, and used the savings on things like video card, HDD, and memory.

AMD's best value days were the Athlon XP times, and post-C2D price drops, imho.

EDIT : The Pentium D 840 was also never more expensive than the X2 4800+, although the stupidly expensive EE 840 was. But the EE was similarly a rape job like the $800-$1000 FX chips, only fit for the ludicrously wealthy and/or stupid folks.

 

[crisium]

Yeah, the Barton 3000+ was a little more than $300 and I remember people complaining about the "most expensive AMD cpu ever". Athlon XP days were nice - AMD's equivalent to P4 always cost less.

 

[zsdersw]

And at what cost? Performance without price is a worthless comparison. You've consistently ignored price since the beginning of this discussion. If the Pentium 4 was 5 times faster than the Athlon 64 but was a million dollars, it would still be a piece of trash. You can't ignore price. Sure, a 3.2Ghz Pentium 4 might beat an Athlon 64 3200+ in a couple tests, but you know what? That doesn't matter at all and doesn't in any way disagree with anything I've ever said. Do you know why? A 3.2Ghz Pentium 4 was more than one hundred dollars more than an Athlon 64 3200+. A 3.8Ghz Pentium 4 would be faster than an Athlon 64 3200+ also. Why aren't you making that comparison? Because it's a lot more expensive. But so is the 3.2. Compare the Athlon 64 to something from the same price.

When a 2.8Ghz Pentium 4 consistently matches or beats an Athlon 64 3200+ and a 3.4Ghz Pentium 4 consistently matches or beats an Athlon 64 4000+ then you can claim the Pentium 4 was competitive because those processors were the same price. The difference increases when you move to dual core. A Pentium D 840(2x3.2Ghz) was MORE expensive than the Athlon 64 X2 4800+(which was two Athlon 64 4000+ processors). At every single price point, the Athlon 64 was unrivaled, which is what I said at the beginning and has always been true.

The only time you find the Pentium 4 winning is if you compare based on performance only which should never be done.

When Anand says something like the Pentium 4 is better at a certain task, he is saying it as if one is comparing a 3.2 P4 to a 3200+ 64.

Faster performance for a higher cost is hardly new. If you used the computer for the tasks in which the P4 was better, the higher cost could be justified. That remains true today with the current offerings from AMD/Intel.

 

[Arkaign]

Faster performance for a higher cost is hardly new. If you used the computer for the tasks in which the P4 was better, the higher cost could be justified. That remains true today with the current offerings from AMD/Intel.

The sad thing about this is that yes, the P4's were more expensive than the Athlon XP counterparts. That all ended during the A64/X2 era, until AM2. The AMD chips, particularly the X2s, were a little to a LOT (often more than double) more expensive than the Intel chips (outside of the stupid EE chips).

 

[Gulzakar]

It doesn't seem like a matter of performance relative to Intel. It really seems to boil down to advertising and market perception, coupled with meeting demand.

 

[Hacp]

Um the Athlon 64 was always an overclocking beast esp the 3000+. Plus the price of AMD mobos were cheaper than Intels.

 

[crisium]

I guess the point is - depending on price point - each company has always had a certain better option. Even today, AMD is great if you want a $99 Quad-Core, for example. A lot of the under $200 market is a wise AMD buy. The executive saying to never buy AMD, especially in 2004 at their height, is retarded.

 

[Edgy]

I think more than anything else it was the volatility of chipset/sockets. You had 940, which required registered, 754 which was EOLed almost before it began, and then 939 which ended up being EOLed too quickly. I mean, great that it had a performance crown, but that was just horrible execution and probably contributes to the validation issue.
.

AMD's issue they ran into when they had better products in the past was not validation complexity due to frequent socket transitions. Validation issue due to socket transition implies that the idiots designing the system did not even bother to check & make sure CPU & MB are same sockets. This is not validation phase, it's critical fail at design phase. Validation processes vary depending on intended use and requirement of finalized product but a typical validation focuses lot on potential undocumented issues rather than well-documented problems such as socket mismatch in CPU/MB components.

Rather, the issue discussed at hand in the article was that customers, large or small, wanted to deal with platform solution provider versus CPU supplier + MB supplier + sometimes graphics chip supplier, and this is not only a matter of convenience but an advantage to the customer (consistency in support, technology, and services) which cannot be easily dismissed.

Many people think that AMD purchased ATi because of the graphics business/gpgpu and as such will eventually birth Fusion technology.

This is true to an extent, but the underlying motivation for ATi purchase was for AMD to mimic Intel to become a platform solution provider because as AMD sales force were finding out at the time, many customers won't even let them in the door even though they had the advantage of having a better regarded CPU product at the time.

As for socket migration management, I personally much prefer AMD's socket to socket (CPU/MB) compatibility with relative simplicity (BIOS update notwithstanding) than Intel's where some existing socket MB may be compatible with new CPU of same socket, sometimes with BIOS update only, sometimes BIOS update only ensures partial list of new CPU compatibility and the kicker... same socket MB which are just completely incompatible with new CPUs...

This was the story with socket 775 and C2D CPUs I believe...

 

[nyker96]

I can't believe this was posted here at AT . Didn't the OP read the XS topic .If ya want something good to talk about Intel posted earnimgs yesterday of Over 2 billion after they paid AMD 1.25 Billion Record earnings for Intel Now thats interesting

I don't see why this post is a problem, it's just news. Maybe a bit on the rumor mill side but it's still something related to Intel/AMD. And if the amd exec did say these things, it's actually true. I'm glad someone there is acknowledging the truth even know it is unpleasant.