Intel bribing THG? Is it possible?

Ncogneto · Sep 27, 2003

Why? Beause the P4EE has a desktop pin configuration, runs in a desktop motherboard, uses desktop RAM, and won't fit into a Xeon motherboard? If that doesn't hold enough water to differentiate it from a Xeon then what does?

Agreed. Or look at it as the p4EE is the barton core p4 same difference. Kudos to Intel. Now get the dam thing on the shelve. Paper launches by anyone is bs. Truth of the matter is the p4EE is just a stopgap for Intel nothing else.

I am confusing marketing with hardware? I seem to be the only one that knows where the line between desktop and workstation hardware is because I refuse to let marketing decide for me. Do you even realize how flawed your statement was?

Give this argument a rest already. Am I not allowed to play games and surf on my workstation? What is the G5?

If that opinion works for you, then that's nice. The rest of the world may not agree, but that's your choice.

As times change so do defintions. The rest of the world is not in agreement with you either.

You can call it anything you want, but that doesn't change what it actually is. Just like AMD can call it an A64FX all they want, but it's still just an Opteron.

Ok then what is a celeron??????

It's not what they tell you, its what they don't tell you!

Crashman · Sep 27, 2003

You have my permission to harsh Intel for the P4EE, as this is an announced product. In fact, it's even a released product as far as I know, it's just not shipping in volume. So this would be a "paper launch" reminicant of AMD's past, or the 1.13 that was delayed (but worse, recalled).

I'm hoping the 939 pin processor will work in 940 pin boards, but to tell you the truth I wouldn't buy an early stage board nor VIA chipset, because of bad experiences.

AMD NEEDED the on-chip memory controller, because it was their only hope to get back into the ring. Look, the A64 3200+ performs good, and is basically what, an XP3000+ with 64-bit extensions?

Only a place as big as the internet could be home to a hero as big as Crashman!
Only a place as big as the internet could be home to an ego as large as Crashman's!

kinney · Sep 27, 2003

Why? Beause the P4EE has a desktop pin configuration, runs in a desktop motherboard, uses desktop RAM, and won't fit into a Xeon motherboard? If that doesn't hold enough water to differentiate it from a Xeon then what does?

Because I've already gone over this with you.
Theres no rule that says those differences between the Xeon and the P4EE is ok.
And that says the difference between a A64FX and a Opteron somehow invalidates it as a desktop processor.

We also went over why desktop processor is nothing more than a processor on a desktop!

The only people that could possibly have the authority to lay such a definition in stone would be AMD or Intel. Each for their respective processors. Even then the boundries would be under constant change.

I seem to be the only one that knows where the line between desktop and workstation hardware is because I refuse to let marketing decide for me.

Actually, you ARE letting marketing decide for you.
Just because AMD markets the Opteron as workstation class hardware doesnt mean that it can only be used as such.

The rest of the world may not agree, but that's your choice.

I would reason that the majority would agree with me. Otherwise I would have never posted a poll.

And until the Socket939 version comes out, that's all it ever will be.

AH-HA! So THATS where you draw the line between server class hardware and desktop class hardware!
One less pin and everything will be ok in your mind?
LOL.

-----
eden is my intel/ati superboy

Groveling_Wyrm · Sep 27, 2003

(Replying to the last post...no reference to author of last post intended)

[hypothetical situation]

I own Ford Motor Company, and have just produced a brand spanking new Ford F150 pickup truck with extended cab. It has a ford truck engine, a ford truck cab, a ford truck bed, ford truck tires, ford truck hubcaps even. Now, is this a ford truck?

However, due to marketing concerns, and the fact that my ford mid-size car sales are inadequate, I am going to reclassify this nice new ford truck as a mid-size econo car!

Now, this new line of F150 econo car/trucks carry up to 5 people comfortably, carries lots of luggage in the trunk/bed, has decent horsepower, and even gets decent gas mileage!

[/hypothetical situation]

As per this example...Now, is my new F150 a car or a truck? As the OWNER of Ford, I call it a car, but as the standard goes...it IS a truck. Who's wrong? Me, as the owner? Or the people, who would recognize this as a truck?

I could call it anything I want, but the fact remains that the basic design of the vehicle was that of a TRUCK. It may work for some of the same things that a mid-size car would, but it is still a truck through and through.

Now, lets take a look at AMD's offering...

It was DESIGNED as a Workstation/Server processor. It takes a Workstation/Server board to run this. AMD can say that it's a desktop processor...but its design says differently...who's wrong?

Would it possibly be better to classify it as a desktop/workstation/server hybrid processor then?

Also, as hard as I have tried, I can not find a single reference to a definitive release date for the P4 EE chip. Doesn't the company officially have to give an official release date, then not release the chip on that date, in order for it to be a paper launch?

As far as I can tell, Intel just released the P4 EE in limited quantities. NO P.R., no major hype build-up (which is typical of Intel), no major announcements. They just said..."Here you go. Here is our new P4 EE". How is this a paper launch?

I have to say that you guys seem to be missing a few OBVIOUS facts, and are stereotyping because of this. After all the different replies and conversations in this thread alone, I keep hearing the terms "fanboy", "Intelliot" and such, way too much. Isn't the majority of the population that posts here at THG, adults?

 Ok, so you have to put your "2 cents" in, but its value is only "A penny's worth". Who gets that extra penny?

pied_piper2004 · Sep 27, 2003

Your analogy is flawed. AMD in this case designed the AMD64 as both a server and desktop processor at the same time. They just released the Opteron first because of low yields.

They know from initial design that it would serve both purposes.

-pied

eden · Sep 27, 2003

AH-HA! So THATS where you draw the line between server class hardware and desktop class hardware!
One less pin and everything will be ok in your mind?

Kinney please read before mocking.

They will remove the 2 extra HT links.

God forbid no one can find the difference except a few, like me and Slvr. The A64FX IS an Opteron core chip. It probably has the same exact die size, thermal characteristics, etc. They rebranded it as A64 FX for desktop/workstation use, since Opteron should've been a dual+ server CPU. Sadly I never got the point of the Opteron 1xx serie. That is just the weirdest money-grabbing way ever yet.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new">This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </A> :lol: Edited by Eden on 09/27/03 00:53 AM.

eden · Sep 27, 2003

Silvr I totally agree with you on the Opteron = A64 FX. Just as you are surprised at how people can't see it, I am too.

Yet people will argue Celeron. Um guys, Celerons ARE different. Less cache, different pins, EVEN REDUCED ASSOCIATIVITY (or cache bitwidth)!

I don't get the logic here anymore. 😱

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new">This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </A> :lol:

pied_piper2004 · Sep 27, 2003

I don’t see how you can’t get it? I don’t get your logic

Opteron -> Athlon 64 FX -> Athlon Desktop and Mobile all in 64 bit!

All purpose chip <- smart design on AMD's part.

For consumers and IT professionals, they’re...

-Fast
-Reliable
-Quiet n’ Cool
-Evolutionary (64 BIT!)
-Available
-Cheap – in most cases (-the FX series)

For AMD
-One design to focus on for enhancements, scaling / ramping
-Cheaper to manufacture one core then several different cores
-If it’s good enough for the IT sector and super computers than it’s surely good enough for your desktop.

What does it matter if it’s server chip or desktop chip? If it kicks’s Intel butt and makes AMD money while providing excellent performance to its target audience, then who cares?

-pied

eden · Sep 27, 2003

On the "Is THG Intel biased" issue, I couldn't help but add this tidbit, as I was looking for some data on the recent article:
<A HREF="http://www.tomshardware.com/cpu/20030923/athlon_64-21.html" target="_new">http://www.tomshardware.com/cpu/20030923/athlon_64-21.html</A>

Go read what they got to say on Bapco Sysmark 2002, and consider its alliance with Intel. (most agree it is using loops in Intel's favor).

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new">This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </A> :lol:

NMDante · Sep 27, 2003

That's not entirely true.

If AMD wanted to market this CPU as an "all purpose chip", why didn't they? Money, of course. If they claimed that the A64 or A64-FX was essentially an Operton, why would any IT manager buy an Operton at a more expensive price?

Also, it won't help AMD focus on one design, since most IT/server functions have inherently different needs than consumer desktops. I'm sure an IT professional wouldn't give a rat's ass about how great Doom3 would run on it's workstations.

Evolutionary? Why, because it runs 32-bit apps? There have been 64-bit processors in servers for years, and software to run them. Again, for consumers, 64-bit computing might be needed, but not in the next few years. When you take that the average consumer uses their computer to surf, manage budgets, or even plays an occasional game, 64-bit is not necessary.

Cheap? The A64-FX costs over $700, without mobo and registered RAM. That's not cheap. The A64-3200 costs over $400, again without mobo. Not to mention that the socket 939 is around the corner in Q1 of 2004, which will mean another upgrade for owners of socket 754 CPUs. That's not cheap.

What does it matter if it's a server chip or desktop chip? AMD's server dept. better make it matter, or IT/server folks will quit buying the higher priced Operton and get the A64-FX, since it pretty much does the same thing.

How many watts does it take to get the center of CPU core?

pied_piper2004 · Sep 27, 2003

That's not entirely true.

If AMD wanted to market this CPU as an "all purpose chip", why didn't they? Money, of course. If they claimed that the A64 or A64-FX was essentially an Operton, why would any IT manager buy an Operton at a more expensive price?

--Well Opteron goes through strict testing and validation, does it not… couple with reliable ECC RAM – but it’s essentially same core is it not – that’s what they are arguing about Opteron = Athlon FX

Also, it won't help AMD focus on one design, since most IT/server functions have inherently different needs than consumer desktops. I'm sure an IT professional wouldn't give a rat's ass about how great Doom3 would run on it's workstations.

--No, why wouldn’t they? Especially if they are designing the damn game Doom 3 (or some other intense work) they need fast and reliable computing power? - See above

Evolutionary? Why, because it runs 32-bit apps? There have been 64-bit processors in servers for years, and software to run them. Again, for consumers, 64-bit computing might be needed, but not in the next few years. When you take that the average consumer uses their computer to surf, manage budgets, or even plays an occasional game, 64-bit is not necessary.

--No because it’s 64 Bit now!!! – don’t speculate on something you cannot possibly predict. I think the whole industry might want the 64 BIT revolution to happen and soon considering people are just not buying computer like they use too! The industry wants the next big thing, it brings new money – AMD might have it here, nothing has happen since the internet. 64 might brings in many new market possibilities.

Cheap? The A64-FX costs over $700, without mobo and registered RAM. That's not cheap. The A64-3200 costs over $400, again without mobo. Not to mention that the socket 939 is around the corner in Q1 of 2004, which will mean another upgrade for owners of socket 754 CPUs. That's not cheap.

--How many times has Intel changed their boards? Live with it. AMD have held out as long as they could… this isn’t just another speed bump chip we are talking about here.

What does it matter if it's a server chip or desktop chip? AMD's server dept. better make it matter, or IT/server folks will quit buying the higher priced Operton and get the A64-FX, since it pretty much does the same thing.

--Price is not an issue for them, reliability is, and with all the super computer wins that AMD is amounting, it looks like it isn’t much of a problem. Nice try though.

How many watts does it take to get the center of CPU core?

Edited by pied_piper2004 on 09/27/03 02:38 AM.

NMDante · Sep 27, 2003

--Well Opteron goes through strict testing and validation, does it not… couple with reliable ECC RAM – but it’s essentially same core is it not – that’s what they are arguing about Opteron = Athlon FX

Operton = server. A64-FX = desktop. That's what AMD wants to market. It's not true, but that's what they want. Using registered ECC RAM is not a consumer desktop friendly item, is it? So what if Operton goes through strict testing and validation, every AMD CPU better go through the same testing/validation.

You never answered my question - Why should a company pay more for essentially the same CPU???

--No, why wouldn’t they? Especially if they are designing the damn game Doom 3 (or some other intense work) they need fast and reliable computing power? - See above

Because most consumer don't compute at a level as high as large companies. Running 32 Office apps isn't going to bog down the server, nor is running 50 Outlook apps. But when did any consumer ever need to run that many apps at one time, and need the same computing power as a server/workstation? They don't, and never will. Not unless they are at home doing their own Earth simulations.

--No because it’s 64 Bit now!!! – don’t speculate on something you cannot possibly predict. I think the whole industry might want the 64 BIT revolution to happen and soon considering people are just not buying computer like they use too! The industry wants the next big thing, it brings new money – AMD might have it here, nothing has happen since the internet. 64 might brings in many new market possibilities.

So, I can't speculate, but you can. "I think the whole industry might want the 64 BIT revolution to happen..." B.S. Every software developer is probably looking at A64 as a test bed to determine if they should start making 64-bit software. If A64 doesn't sell to large market share, why would companies spend the money to appease a small market? Why spend millions, to get back thousands? Nothing has happened since the internet? For who? Consumers? Professionals? Plenty has happened. 300mm processes. SATA H/Ds. Higher DDR memory speeds. It depends on what you look at.

--How many times has Intel changed their boards? Live with it. AMD have held out as long as they could… this isn’t just another speed bump chip we are talking about here.

How many times did intel launch with 2 different sockets, and announce a third before launch? Hey, it is life, but it's something AMD had the time to fix, prior to launch.

--Price is not an issue for them, reliability is, and with all the super computer wins that AMD is amounting, it looks like it isn’t much of a problem. Nice try though.

Nope. Sorry, but price is a major issue for IT/Server departments. Super computer wins are great, but how much of a discounted price are the CPUs being sold at. The China super computer didn't even disclose the costs of the system. It's good for AMD to get some share of the workstation/server market, but without a healthy consumer base (which price is an object) AMD won't make too much of a profit. Plus, if anyone who wanted to start a server with little money chooses the A64 line over the Operton line, I bet AMD will start to worry.

How many watts does it take to get the center of CPU core?

kinney · Sep 27, 2003

This isn't all direct at you. I responded simply to some of your points and to anyone involved in the conversation..

Your hypothetical situation is different from the one here.
I don't know if you meant it to be.

Websters definition of a truck-
" f : an automotive vehicle with a short chassis equipped with a swivel for attaching a trailer and used especially for the highway hauling of freight; also : a truck with attached trailer"

Your ford truck fits that description perfectly. And if I have to define each word in mid-size econo car to prove that a ford truck is not one, I will.

Now, websters definition of a server-
"6 : a computer in a network that is used to provide services (as access to files or shared peripherals or the routing of E-mail) to other computers in the network "

And a workstation-
"2 a : an intelligent terminal or personal computer usually connected to a computer network b : a powerful microcomputer used especially for scientific or engineering work "

I could not find any hardware specifications for any of the keywords used in this thread.
Hence it is based OPINION and the individuals perception of the marketing.
It would have been much easier for slvr to agree with me on this fact (that it ends up being his opinion vs mine) when I brought it up a few respones ago.

Would it possibly be better to classify it as a desktop/workstation/server hybrid processor then?

Unfortunately there is no such thing. They are all CPUs. Some work faster than others, thats it when it comes down to it.
Desktop/workstation/server are marketing terms for present day hardware. They are simply references to the consumer, if AMD choses to reference an Opteron dressed in A64FX clothing as a desktop, thats what it becomes.

I think the idea that hardware class is defined by the actual hardware itself has been fully debunked...
You can use "server hardware" or "desktop hardware" interchangably in the PC environment.

Tell me a piece of software that WILL NOT run on a desktop machine in the PC environment, esp now that desktop and workstation have infinitely blended FURTHER with "desktop" PCs now incorporating 64bit.

-----
eden is my intel/ati superboy

kinney · Sep 27, 2003

I agree the A64FX is an Opteron based CPU. Just rebadged, thats it. Finito. I've never doubted that fact.
Now would be a good time to remind everyone the P4EE is a Xeon MP based CPU.

And?
Does that change the fact that either one of those can be used as a server, as a workstation, as a desktop?
The opterons a great desktop cpu. Just as a P2 was a great server cpu back in the day.
Just as the IBM deep blue (just used for example) I use in my spare bedroom is a great workstation processor.
If you can afford whatever people assume is meant for a certain task use it for a desktop then more power to them. Because no boundries, software or hardware will stop them in the PC environment.

And if people (yourself and slvr), are assuming that the marketing terms being thrown around are so entirely true (because that is the only weak basis you have for differentiating between server and desktop CPUs), that when somehow, in their own mind, assumed AMD's Opteron cannot be desktop hardware..

If that is your mentality, then what in Gods name would stop them from believing that when AMD does tell us it has become desktop hardware... THAT IT ACTUALLY IS.

So what, it works one way, that you believe in following differences set through marketing naming schemes between server and desktop hardware.. But when AMD claims the Opteron is now the A64FX ("desktop hardware"), you (or at least svlrphoenix) cried foul.

Sounds like bias creeping in to me because you can't follow one part of that equation and not the other half, which just so happens to be friendly to AMD.
Surprise surprise again.

-----
eden is my intel/ati superboy

pied_piper2004 · Sep 27, 2003

Operton = server. A64-FX = desktop. That's what AMD wants to market. It's not true, but that's what they want. Using registered ECC RAM is not a consumer desktop friendly item, is it? So what if Operton goes through strict testing and validation, every AMD CPU better go through the same testing/validation.

You never answered my question - Why should a company pay more for essentially the same CPU???

--Read again, Operon gets more rigorous testing then the Athlon line, for reasons obvious. No different than Xeon and P4. That line is blurring anyways, as we move toward multi media network streaming video / audio home base centers.

--Besides with the extra HT links, Opterons are meant to scale, 4 way to 8 way and so forth, sure your gonna pay extra for that.

Because most consumer don't compute at a level as high as large companies. Running 32 Office apps isn't going to bog down the server, nor is running 50 Outlook apps. But when did any consumer ever need to run that many apps at one time, and need the same computing power as a server/workstation? They don't, and never will. Not unless they are at home doing their own Earth simulations.

--Stop down playing it. That line is blurring more and more every day. We are moving into a new digital age. Many people are demanding more and more from their home systems. AMD is providing future needs now!!! You’re missing the whole point behind the AMD 64! – That’s the same mentality as “we will never need more than 640k”

So, I can't speculate, but you can. "I think the whole industry might want the 64 BIT revolution to happen..." B.S. Every software developer is probably looking at A64 as a test bed to determine if they should start making 64-bit software. If A64 doesn't sell to large market share, why would companies spend the money to appease a small market? Why spend millions, to get back thousands? Nothing has happened since the internet? For who? Consumers? Professionals? Plenty has happened. 300mm processes. SATA H/Ds. Higher DDR memory speeds. It depends on what you look at.

--The industry IMHO has more to gain from this other then just sitting on old technology. 64 bit is imminent. It will open new possibilities and with new possibilities brings new markets and with new markets, bingo you guessed it, more money. If your arguing that AMD might be too early, then perhaps that’s true. But any way you slice it, 64 will be the next wave.

--again you go off with more speculations.

How many times did intel launch with 2 different sockets, and announce a third before launch? Hey, it is life, but it's something AMD had the time to fix, prior to launch.

--How many times did Intel screw their customers? Let’s not get into a pissing contest. AMD has a far better record then Intel, and they have always been a lower cost solution. I’ll never forgive Intel for what they were charging for a Pentium 1. You should be thankful that AMD came along.

--Besides, look how Intel stagnate, they keep us wanting to pay for every speed bump. How long would they have held on to the P4c or 32 bit on the desktop if Athlon64 wasn’t produced? Look who comes out with an emergency edition, out of the blue, forcing review sites to include in the AMD benchies. AMD is truly giving us the future.

Nope. Sorry, but price is a major issue for IT/Server departments. Super computer wins are great, but how much of a discounted price are the CPUs being sold at. The China super computer didn't even disclose the costs of the system. It's good for AMD to get some share of the workstation/server market, but without a healthy consumer base (which price is an object) AMD won't make too much of a profit. Plus, if anyone who wanted to start a server with little money chooses the A64 line over the Operton line, I bet AMD will start to worry.

--Nope your wrong, Xeon don’t cost much more to make than P4s. Opteron doesn’t cost much more the Athlon. See where I’m going? Both companies charge more, why do you think? Is it because the IT/Server departments are willing to pay more? It’s all about responsibility. The IT manger would be killed if he had bought and installed a desktop CPU for a mission critical server. After all that’s why they pay the big bucks!

--But they are all the same core, essentially. Of course some specific features are built into server chips (i.e. remote monitoring)…

With the extra validations, the HT links, Registered RAM ECC and the scalability of Opteron vs. the desktop counterpart, that is enough to pay the premium.

Edited by pied_piper2004 on 09/27/03 04:12 AM.

Ncogneto · Sep 27, 2003

Yet people will argue Celeron. Um guys, Celerons ARE different. Less cache, different pins, EVEN REDUCED ASSOCIATIVITY (or cache bitwidth)!

Isn't the celey just a rebadged P4 with a Willy core?

It's not what they tell you, its what they don't tell you!

saintbabs · Sep 29, 2003

AMD is stuck between a rock and a hard place: it can only produce variations of the same AMD64 architecture, which is a very nice argument in the server, and certain workstation market segments.

Whatever AMD does on the desktop to justify this AMD64 architecture, cut-down Opteron, or overpriced Athlon, however you look at it, is a hard sell.

AMD can't drive the price of the Athlon 64 down because, that belittles the arguments made in its favor, yet, they expect to sell it on the basis of a future shift in applications and Windows that is very uncertain.

I don't want to hear about Linux. AMD has put all its desktop 64-bit hopes into XP 64-bits from Microsoft. It's very clear in the marketing.

So, you have a bunch of mixed messages, and a strategy built around one feature, 64-bit extensions.

It just doesn't work.

It's not even clean, or simple.

It's confusing. It's not aimed at a particular user, but it has the fanboys excited, but since when are the fanboys going to change the minds of the average consumer.

They belittle everything that goes againts their contrarian views so, Athlon 64 is great marketing for geeks who already believe, but it does nothing to convince the skeptical, or the fence-sitters, or the people who don't really give a damn as long as they get a decent work PC for a decent price.

Omid Rahmat
GM & Publisher
Tom's Guides Publishing
www.tomshardware.com

Incitatus · Sep 29, 2003

Wow are these really the kind of people that work at THG? People that have draw conclusions even several months before any real developments have a chance to reveal themselves.

Edited by Incitatus on 09/29/03 02:12 PM.

eden · Sep 29, 2003

Now would be a good time to remind everyone the P4EE is a Xeon MP based CPU.

And vice-versa, a Xeon MP is a P4 in disguise too.
What's your point? We could go down to the core itself and argue and easily agree that K8 is K8, P7 is P7. Our goal here was merely to prove that the K8 Athlon 64 FX is the exact chip used for servers, and that selling it as that, especially when there is a uni-processor Opteron version that's the same thing all over, shows a lack of respect towards consumers. It only screws those who buy the 1xx Opteron. Pretty soon some enthusiast mobo maker will likely make Athlon 64 FX compatible boards to ride on the Opteron dual mobos, if not just find a pin tweak. What does that make for AMD? Excessive use of a useless chip known as the Opteron 1xx.

And if people (yourself and slvr), are assuming that the marketing terms being thrown around are so entirely true (because that is the only weak basis you have for differentiating between server and desktop CPUs), that when somehow, in their own mind, assumed AMD's Opteron cannot be desktop hardware..

If that is your mentality, then what in Gods name would stop them from believing that when AMD does tell us it has become desktop hardware... THAT IT ACTUALLY IS.

So what, it works one way, that you believe in following differences set through marketing naming schemes between server and desktop hardware.. But when AMD claims the Opteron is now the A64FX ("desktop hardware"), you (or at least svlrphoenix) cried foul.

Oh I said that?
For real? I debated server vs desktop? PROVE ME. Prove me ya smarty-pants, go ahead! Prove me that I argued that. It seems you're also out to get us for the workstation thing. US, as in ME and Slvr. Where ME stands in there, I don't know.
<A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=36513#36513" target="_new">http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=36513#36513</A>
Well what do we have here. I sure hope you didn't forget that I was in agreement hmm?

To me, a server CPU can be any one out there, however the distinguishable ones, i.e. those that are sold as that, in case you did not notice, come with different specs both internally and externally. And with server capabilities, Opteron chips have the best proof out there with the multiple HT links, that they are made for servers. Therefore, you CAN make an A64FX a server chip, because basically, you can put it in Opteron mobos and it'd be the very same multi-CPU system. Of course you can make a server with any CPU you want, but specifically designed ones like the Opterons, make sure you can have a non-uniform memory adressing, further making the point that they are made for full-blown server use.

Sounds like bias creeping in to me because you can't follow one part of that equation and not the other half, which just so happens to be friendly to AMD.
Surprise surprise again.

Bias towards who? My arse? Or did it come out of yours?

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new">This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </A> :lol:

spud · Sep 29, 2003

Well said well said, but like I always say fagboy fanboys don’t see it this way and never will. They see it as simply as AMD is a visionary company that makes CPU's pieces of silicon that rock. They see the CPU as a substitute to their penis and the company as their egos.

The follow blindly and that’s what the problem is AMD actually believes these misfits that are trying to break from the norm will deliver them from the red into the ranks of IBM and Intel in sales power money and influence. It’s really sad that this seems to be happening but it's the same in any industry... look at the automotive junkies.

-Jeremy

:evil:

<A HREF="http://service.futuremark.com/compare?2k1=7013108" target="_new">Busting Sh@t Up!!!</A> :evil:

<A HREF="http://service.futuremark.com/compare?2k3=1311896" target="_new">Busting More Sh@t Up!!!</A> :evil:

CorwinB · Sep 29, 2003

Wow are these really the kind of people that work at THG? People that have draw conclusions even several months before any real developments have a chance to reveal themselves.

Realistically, how long do you expect people to give a free pass for AMD to hype Athlon64 as the second coming ? It's not even "the second come-back"...

Of course, AMD fanboys are all over those forums spewing random (and totally expected) nonsense about how Intel "bought THG", whereas they used to consider these boards the last place of free thought or some similar stupidity when THG ran the stories about Rambus performance or when their findings forced Intel to recall the 1.13GHz P3. And back then, of course, the Intel fanboys were ranting about how Tom was on AMD's payroll...

Regarding the issues at hand :

Face it, the A64 is not another Athlon. It looks like a decent and promising chip, whose sales potential is held back by ridiculous design/marketing decisions (ECC-RAM anyone ?). But it's not enough to even bother Intel past the release of the P4EE (heck, even the original Athlon didn't change the perception of Intel's chips, and it creamed everything Intel could try to get out of the P3 core). And the Inquirer can claim they heard a lot of things about MS forcing Intel to go with AMD 64 bit extensions, that's a plain, simple load of BS, as there are already very nice and powerful 64 versions of MS OSes that run pretty well on Intel's current 64bits line, thank you. Of course, it's no use trying to get that in a fanboy's head. I'll gladly give Omid some style points for trying in a no-nonsense, no-holding back way, though.

Regarding the remarketing of the Opteron as a "desktop" chip, that has to be one of the most stupid, insane, and self-destructive decisions ever made by a company in this field. When Intel tried to force Rambus down people's throats, many were rightfuly angered because the product didn't perform very well, and was extremely expensive. Now, of course, since a similar decision (minus the "closed standard" part, I agree) is coming from AMD, some of the same people find it's an excellent idea... People thought the Rambus idea because they wanted to be able to keep at least their expensive memory chips when they moved from a format to another (even at the cost of slightly reduced performance). Now, if you want to go the Athlon64 FX road, you have to get an expensive chip, an expensive motherboard, and super-expensive RAM. Ouch. People already owning recent Intel boards and DDR400 RAM(ie true performance enthusiasts, not "cost-driven" enthusiasts or "mindless AMD fanboys") can easily upgrade to a P4EE and get similar performance for a fraction of the price of switching to an AMD solution. So AMD can't reach its traditional cost-driven segment (its new processors and especially the surrounding platforma are way too expensive for cost-driven business), nor companies (face it fanboys, most IT directors will laugh in your face if you suggest AMD, and not all of them (in fact probably none of them) are on Intel payroll, nor are they totally clueless). They are left with their core audience of fanboys, and will probably drive part of the "money is no object" performance fans. They will get a few months under spotlight from favorable reviews, and perhaps some mindshare, but unless they drastically alter their marketing, I don't see them in business anymore in a few years, and that would be a damn shame, both because the company made some extremely good chips, and because it would force us back to the old Intel pricing policy.

Ncogneto · Sep 29, 2003

Whatever AMD does on the desktop to justify this AMD64 architecture, cut-down Opteron, or overpriced Athlon, however you look at it, is a hard sell.

And just exaclty why is that? It has proven to be faster in most apps then the p4c 3.2 and arguably as fast as the p4EE 3.2 (unless you choose to run multiple benches of quake at 640 x 480 to stack the results). I speculate less than 1% of your readership could careless that it is architually the same as a server chip for whatever reason. Fact of the matter is who cares? Does it get the job done? Is it price competitive with the intel offerings? The answer is a simple yes.

So, you have a bunch of mixed messages, and a strategy built around one feature, 64-bit extensions.

It just doesn't work.

I tend to disagree. Itanium would be a strategy built around one feature much more so. 64 bit extensions is a marketing strategy much like intel used there "netburst". Face it, AMD has it, Intel doesn't it. If the roles were reversed do you think Intel would flaunt it as well? Of course they would.

It just doesn't work.

If you are to make statements like that would you please extrapolate a little bit as to why?

It's confusing. It's not aimed at a particular user, but it has the fanboys excited, but since when are the fanboys going to change the minds of the average consumer.

So yes, AMD be damned for designing a processor that can perform admirably across platforms....is this what you are saying?

They belittle everything that goes againts their contrarian views so, Athlon 64 is great marketing for geeks who already believe, but it does nothing to convince the skeptical, or the fence-sitters, or the people who don't really give a damn as long as they get a decent work PC for a decent price.

Ok, and yet you say THG is not biased? You do realize you have just insulted close to half your readership. But thats ok you still have die hard users like spud on your side that give us such valuable insight as the post above.

It's not what they tell you, its what they don't tell you!

Ncogneto · Sep 29, 2003

Just exactly what is AMD ramming down our throats? You seem to be missing a few key points.

Athlon64 fx. Yes, you have to get a new motherboard and RAM. And this is different than past processors in what way? How long have we had socket A? Through that time how many board changes did Intel go through? Has AMD stood up and said, "sorry folks we are no longer supporting unbuffered DDR in any future proccesors"? Nope. This is in fact what Intel did when we first were introduced to RDRAM. Hmmm, does AMD have any incentives with makers of ECC ram like Intel did with RAMBUS ? Nope. Give it a rest. Ok, so you don't want to spend the exta money on ECC ram? Great, all you have to do is wait until the 939 pin version comes out. Hell, by that time maybe you will actually be able to buy a p4EE.

But it's not enough to even bother Intel past the release of the P4EE

Fact of the matter is if it wasn't a bother to Intel they would have never released (I use the term released lightly) the p4EE. Your argument is baseless. In one hand you critisize AMD for releasing a server chips with albiet minor alterations and market it as a desktop chip and at the same time acknowledge Intel has done the exact same thing. People in glass houses should not throw stones my friend.

It's not what they tell you, its what they don't tell you!

sonoran · Sep 29, 2003

>>>Itanium would be a strategy built around one feature much more so.

Itanium:
264 Application Registers
64 Predicate Registers
11 Issue Ports
6 Integer execution units
3 Branch execution units
2 FP execution units
1 SIMD execution unit
2 Load and 2 Store execution units
Up to 6MB On-die cache
6 instructions/cycle

Opteron:
40 Registers
6 Issue Ports
3 Integer execution units
Fmisc execution unit
Fmul execution unit
Fadd execution unit
1 SIMD execution unit
2 Load or 2 Store execution units
Up to 1MB On-die cache
3 instructions/cycle

"One" feature? That's a bit disingenuous, isn't it?

saintbabs · Sep 30, 2003

This is where I just get tired of all the pro-AMD64 sentiment. It seems like AMD64 is whatever you want it to be at any time you want it to be. Yet, it can't possibly be all things to all me because, and listen carefully, it doesn't have the resources. It is spread thin. It needs focus.

Can you compare an Opteron or Athlon 64 to an Itanium? They're not even playing in the same ballpark. Intel is targeting a different segment of the market, and frankly, you could just as easily argue that by building a 64-bit architecture from the ground up, Intel is doing the computing world a favor by trying to move us away from x86 architectures. If 64-bit has so many advantages, why not just build a true 64-bit architecture?

But, I don't care about that. It's a pointless argument when talking about Athlon 64s, and dubious when talking about Opterons. They are designed to be stop-gap products when the gap is not even defined, or apparent. The gap between 32-bit and 64-bit computing is a marketing induced gap. It's not one that has been created by demand outstripping supply, which is what should have happened.

Opterons compete with Intel Xeons. I have said that AMD has a clear advantage over Intel that it should be exploiting. I am not sure that they have the resources to do that.

As far as Athlon 64s are concerned, how long before there is a viable 64-bit software market that justifies the AMD feature set for 32/64 bit migration? A year, two years, three years? If there is a momentus move to 64-bit apps because, suddenly, everyone goes XP 64-bit crazy then, what, Intel is going to suddenly say, "Whoa, we lose."

How difficult does anyone think it is going to be for Intel to come up with 64-bit extensions and get software support for it, no matter how long of a lead AMD has?

At the end of the day, for the next couple of years, the market is going to be so overwhelmingly 32-bit that the 64-bit feature set is not a compelling argument for anyone to switch to AMD.

AMD enthusiasts will talk it up, but they haven't managed to get AMD a profitable quarter in a long time, and that's all that matters: profitability.

Technology leadership and innovation is immaterial in this instance. AMD has done some good work from an engineering perspective, but the computing world is littered with the bodies of companies that did good engineering only to have it fall apart in sales and marketing.

Yet, I don't see anyone give me a compelling argument, or showing me how AMD has got its act together on the sales and marketing front.

All everyone says is they need to get a Tier One OEM. Oh, yeah, they have IBM, and they have Sun. They've had HP and Compaq, in the past, but at what cost to their bottom line?

There is no strategy for AMD64 beyond build it and they will come, and I don't see why AMD enthuasists should be responsible for paying upfront to keep AMD in the game for the future.

It used to be that early adoption was necessary because, people couldn't get a hold of enough PC performance fast enough. Compaq built its business on being the fasttest PC, and the first to market. That doesn't matter anymore.

In fact, Dell will take 6 months before it actually catches up with the market at the high-end, preferring to wait for all the bugs to be ironed out.

At the end of the day, all I am seeing is regurgitated AMD PR, and while I understand everyone's desire to see AMD succeed, or be strong, or be competitive, this is capitalism, and the strongest survive.

If AMD can't find the strength then, we should let it find its own way instead of propping it up.

The problem is that with the enthusiast community behind it, AMD doesn't have to make any hard decisions. It falls back on its fan base, and that, to me, and this is my opinon only, is a sign of weakness.

It's a philosophical argument.

Omid Rahmat
GM & Publisher
Tom's Guides Publishing
www.tomshardware.com

Intel bribing THG? Is it possible?

Distinguished

Polypheme

Distinguished

Distinguished

Distinguished

Champion

Champion

Distinguished

Champion

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Champion

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Distinguished

Share this page