Thoughts on what's happening with AMD vs. Intel

[Pretender]

To start off, I don't want this to be a flame war of "Intel Rulez", "AMD is tha best". A civilized debate would be nice, but that's probably just too much to ask so I'll just ask for your thoughts.


With their P4's looking like they're going to be either slower than or as fast as the P3s so far, and the PIII 1.13 ghz's recalled and postponed until next spring, it seems as if Intel is heading down the wrong track, and it seems to have happened rather suddenly. The question is: Where do you think Intel started to go wrong?

 

[apoppin]

At the beginning of this year. After AMD came out with their Athlon.

 

[LXi]

Huh... first of all, the Athlon came out last summer, yes, the summer of 1999. And thats when things started to go wrong for Intel, i820, MTH, RDRAM....

 

[Dark4ng3l]

I think intel wil do like 3dfx but the big diference is that thell just refuse to die out(if they suack like cyrix in the future) because they have too much money.

 

[apoppin]

It's true, LXi, AMD's Athlon came out last year. However, it took awhile for it to sink in for Intel. This is the year of their big screw-ups.

 

[LXi]

Their i820 and MTH, which came out Fall 1999, has been a screw up since day 1, that was basically the lead off screw up for everything else.

 

[Quickfingerz]

I think Intel is trying too hard to compete whereas AMD is going along in their merry little way up the CPU market hill. Only one can be at the top and the way I see it AMD has a decent chance of making it there.

 

[JellyBaby]

Intel's troubles started shortly after introducing the first AGP specification. At right about that time they started trying to dominate more than CPUs and chipsets...remember the i740 (?) video adapter? This truly sucked and they could never compete with true 3D solutions. Shoulda learned their lesson then (to stick with what they're good at). But the biggest damage was done when they went to bed with Rambust Inc. to try to dominate the memory market.

 

[beat mania]

Rambus is Intel's undoing. At least Intel's trying to do DDR now ... even though they're continuing to support Rambus.

And Intel wastes too much resources into those 'value pc' thingie - look at Timna, they had to kill that too. What a waste of time and money. Just stick with crippled P3's already.

And Intel Just loves to put integrated graphics into its chipset, more of that value pc syndrome.

A bit off track, but if the fastest available P4 can be faster than the fastest Athlon then I don't see the problem of putting out a P4. Even if they have to depend on very high frequecy for that.

Anyway, if AMD's flash sales eats into Intel's then Intel'll have a real big problem. Not that it isn't big now.

 

[apoppin]

You're being really picky, LXi.

This is the year Intel's prestige is slipping and they are starting to lose in the market. 1999 was still a banner year for them. Although MTH was released in late 1999, they did not find out about it's problems and do their big expensive recall until THIS year.

Rambus was always poor but they had high hopes for it in 1999 as dominating the RAM market. THIS year they are moving to an expensive divorce.

Last year their chipsets were riding high with high hopes for a real BX replacement. THIS year their hopes were completely dashed.

LAST year they thought they could stay ahead of the Athlon with their P3 Coppermine. THIS year sees all the paper launches and Intel falling further behind schedule. THIS year the Duron completely shut down their Celeron. This year the Coppermine has been dominated by the T-bird especially to clock speed.

 

[LXi]

Im not being picky, Im just being more accurate.

Yes, I agree most of the things happened in the year 2000, but when you look back, they made the decision in 1999 to go with RDRAM, they made the wrong decision to make MTH, they planned everything before the fiasco took place. It wasnt just AMD who took advantage of this disaster, VIA was the main reason they lost a major battle in 1999, the VIA 133A sold big and the i820 quickly faded. It wasnt until this year when the winner and losers are determined. The made the wrong decisions and they eventually led to this. Think about it, if they released the i815E sooner, or at least a more refined BX chipset, they would've skipped all the RDRAM slump which is what should be blamed for.

 

[Vegito]

Agreed... preety much know Intel will slowly sink unless, they stop shoving crap down consumer's throat. Especiall with scambus. but one problem. Currently generation software wont run well for a P4, but with all the microsoft stuff optimized for intels' product, of course P4 runs slow on old software...

 

[beat mania]

One thing, AMD doesn't seem to be as concerned about Duron cutting Athlon's sales, while Intel obviously do not want Celeron to eat into the P3s'? I mean, they're still keeping the Cel at 66MHz, for crying out loud. This probably caused Intel to lose the cheap pc market to AMD, that the Duron is more powerful then the Celeron.

 

[apoppin]

Actually to be completely accurate, Intel's wrong decisions were made at least a year before the actual products were released. So the seeds for their big mistakes can probably be dated back to 1998 or earlier.

However, it is the competition with AMD that really clouded Intel's corporate judgement and caused them to rush products to market before they were ready. Intel never had to hurry products before this year.

 

[LXi]

Ok, their failure is a combinations of a variety of things, but IMHO, the most important and most devistating blow was the RDRAM/MTH fiasco.

 

[DDad]

Intel started having problems about 5 years ago IMO- with the IA-64
After that- the I 740 video flopped big time
It became obvious last year(3rd quarter to be exact)- Coppermine was delayed, the initial I820 boards had to be scrapped 3 days before introduction (and then had to be recalled 6 months later!), there have been chronic shortages of higher speed Cumines, and the growing belief that Intel is more concerned with Image and PR rather than solid engineering- witness the "paper releases" of the assorted processors, culminating in the "paper release" (and subsequent recall) of the 1.13

 

[PliotronX]

I just have a question... why is Intel doing what they're doing?

 

[beat mania]

Actually I think the IA64 is impressive technologically. We don't know how it would be marketed yet though.

Intel decision makers are just not used to these pressures and rush things, plus start up A LOT of projects to try to cover a lot of ground, but that most eventually gets killed. And this thins out the resources, thus the parts that actually gets completed aren't as polished as they should be.

 

[Varborta]

AMD all the way, more advanced CPU. Than old aging pIII cpu.

 

[Pretender]

Plz don't follow Varborta's example. We're trying to have a discussion here.


IMHO, I think one of the main problems is that Intel's engineering dept. seems to be on some sort of illegal narcotic and can't seem to get things right in any quick amount of time. So in order to keep up with AMD, they're either rushing products out the door without making sure they work and without actually having any available to buy.

As for the paper releases, I don't really know the history of these pre-PIII. Were they up to these tactics back in the Pentium/PII/Celery days, or is this a new tactic to try to save face with the current trouble they're having?

 

[apoppin]

In the "old" days, Intel's roadmap was pretty accurate. They were the top dog company and dictated the market. A product came out at their leisure. It was never rushed - they didn't have to rush. They took their time and generally their products were very well researched, developed and tested before release.

This all changed with the AMD competition.

 

[KarsinTheHutt]

i820 and Rambus was the start of the slide downhill. Intel is so diversified that the core chip buisiness didn't get enough attention.

 

[beat mania]

<< IMHO, I think one of the main problems is that Intel's engineering dept. seems to be on some sort of illegal
narcotic and can't seem to get things right in any quick amount of time. >>



Pretender, I don't think any engineering department can do things rushed. Having to rush would inevitably mean missing something. If you skip a few tests to save 2 weeks to meet deadline, then you're compromising the quality of the work.

 

[Midnight Rambler]

People just don't seem to get it, as far as Intel's future.

Yes, they have had a myriad of problems this year. But &quot;go the way of 3dfx&quot;. Put that crack pipe away ...

Intel is moving away from the desktop CPU business. It's becoming a commodity item (just like what happened to DRAM memory in the 90's), and the profit margins are rapidly shriveling for both Intel and AMD when it comes to desktop CPUs.

Intel's focus is clearly on 'Net appliances, servers, hosting, etc. - in general, Internet infrastructure. Add to this StrongARM, a DSP co-developed with Analog Devices, and their strategic partnerships with Ericsson, Cisco, and several other teleco and infrastructure companies. This is where Intel is heading for the future. And with 5-6 additional fabs coming online in 2001, they will regain their momentum, just in different sectors of the overall silicon market.

 

[LXi]

What you're saying is that they dont care about the consumer segment anymore? Dude, they're controlling more than 75% of the PC processor market, why would they want to move away from that?