Intel x86-64. **Updated 2/17** It IS AMD64 compatible...out next quarter

[NFS4]

http://www.anandtech.com/news/shownews.html?i=21418

What are the chances that Intel will licence AMD64 technology seeing as the support and install base is already there? Doing so would be a major marketing coup for AMD...

Anyway, this article made me start thinking about it:

http://www.theinquirer.net/?article=14038


**Updated 2/17** It IS AMD64 compatible...out next quarter

http://news.com.com/2100-1006_3-5160169.html?tag=nefd_top

 

[DAPUNISHER]

There are some pretty bold assertions in that article, but if he's correct then AMD will have something that will make it worth spending the $$$ on a TV advertising blitz to help garner mind share of the average consumer. I recall some of the Intel employees who post here saying that AMD and Intel always cross-license everything so they should be good to go on using AMD's 64bit tech right?

 

[NFS4]

Originally posted by: DAPUNISHER
There are some pretty bold assertions in that article, but if he's correct then AMD will have something that will make it worth spending the $$$ on a TV advertising blitz to help garner mind share of the average consumer. I recall some of the Intel employees who post here saying that AMD and Intel always cross-license everything so they should be good to go on using AMD's 64bit tech right?

Yeah, AMD uses SSE/SSE2 so Intel should be cool with AMD64 AFAIK

 

[OddTSi]

Yup, Intel saw the performance improvements from the A64 about 6 months ago and they decided to build an x86-64 processor themselves. This processor should be ready for demonstration in a week or two.

Anyone see the SERIOUSLY flawed logic in there? If Intel is capable of going from nothing more than an idea to a fully working die in 7 months, they must've invented a time machine or something. Come on guys, let's use that big wrinkly lump of tissue in your head for a second, can we? If Intel is ready to demonstrate a 64-bit Xeon processor in a week then that means that the decision to make this processor was made at the least back when the Athlon 64 was nothing more than a hint on a roadmap.

This assertion by AMD-fanboys (The Inquirer being the biggest of these) that this processor is a direct result of "Intel seeing how much they're getting their butt kicked by the A64" is the stupidest thing I've heard in a while. IF Intel did decide to build a 64-bit processor as a direct result of that and nothing else, we won't be seeing that processor until at least 2 years from now, not less than 2 WEEKS from now.

The ignorance of fanboyism really astounds me sometimes.

 

[DAPUNISHER]

Originally posted by: OddTSi
Yup, Intel saw the performance improvements from the A64 about 6 months ago and they decided to build an x86-64 processor themselves. This processor should be ready for demonstration in a week or two.

Anyone see the SERIOUSLY flawed logic in there? If Intel is capable of going from nothing more than an idea to a fully working die in 7 months, they must've invented a time machine or something. Come on guys, let's use that big wrinkly lump of tissue in your head for a second, can we? If Intel is ready to release a 64-bit Xeon processor in a week then that means that the decision to make this processor was made at the least back when the Athlon 64 was nothing more than a hint on a roadmap.

The ignorance of fanboyism really astounds me sometimes.

Who said anything about 6 months other than yourself? The way I've heard it, Intel and AMD cross-license each other's tech basically immediately upon it's developement, and if so means they can work on integrating it on their own microprocessors long before either have a product ready for sale. How else does AMD manage to incorporate features like SSE2 into a CPU under design for a minimum of 3yrs? from what I've read.

 

[buleyb]

Agreed, this wasn't some last minute job, this was all planned, regardless of timing.

What I would like to see, but don't expect, would be a desktop IA64 chip (only because IA64 is such a better arch design), but that has about 0% chance of happening.

 

[mikecel79]

Originally posted by: buleyb
Agreed, this wasn't some last minute job, this was all planned, regardless of timing.

What I would like to see, but don't expect, would be a desktop IA64 chip (only because IA64 is such a better arch design), but that has about 0% chance of happening.

This would be huge and it's what I would like to see but like you said it has about no chance of happening! Or an Itanium that runs x86 code as fast as a new Xeon. I think if they really pushed the price of Itanium down we would see many more applications available for it now.

 

[OddTSi]

Originally posted by: DAPUNISHER
Who said anything about 6 months other than yourself? The way I've heard it, Intel and AMD cross-license each other's tech basically immediately upon it's developement, and if so means they can work on integrating it on their own microprocessors long before either have a product ready for sale. How else does AMD manage to incorporate features like SSE2 into a CPU under design for a minimum of 3yrs? from what I've read.

I didn't make it clear in my post originally, so I added to it to make it more clear. Everyone and their grandma is saying that Intel made this chip because the A64 is performing so well, etc., etc. That cannot be further from the truth. If this chip is trully going to be demonstrated in a couple of weeks then Intel decided to make this long before AMD had any working silicon of the A64. Long before they could do more than just guess at its performance.

Originally posted by: buleyb
Agreed, this wasn't some last minute job, this was all planned, regardless of timing.

What I would like to see, but don't expect, would be a desktop IA64 chip (only because IA64 is such a better arch design), but that has about 0% chance of happening.

I completely agree. I've been saying for some time now that we need to drop x86 already, even if it means losing hardware backwards compatibility. And if I were to choose an ISA to replace x86 with, I couldn't think of a better one (technically speaking) than IA64. But like you, I also don't think it's likely to happen. I'll be thrilled if it does, but I'm not going to hold my breath for it.

 

[DAPUNISHER]

Ah, I see your edit now, and I agree that it's foolish to believe it's a reaction to AMD's recent success with the A64. Intel is definitely able to develope technologies that will never see the light of day thanks to their financial resources, so if they did develope a 64bit CPU that incorporates AMD's technology they stared the project years ago certainly. The only reason for going public with it that seems sensible would be that they are indeed being pressured by major customers like Dell to provide 64bit desktop goodness soon.

 

[Alkaline5]

Originally posted by: OddTSi
The ignorance of fanboyism really astounds me sometimes.

IMO, the interesting issue here is the fact that Intel has been publically claiming that x86-64 was a waste of time ever since it was announced. Then they turn around and apparently have been working on their own x86-64 processors for years just in case the architecture actually caught on. That's good business strategy (berate the competitor and then cash in on their success at the last second) but they're not going to win any Nobel prizes by talking out of both sides of their mouths.

 

[NFS4]

Originally posted by: OddTSi

I didn't make it clear in my post originally, so I added to it to make it more clear. Everyone and their grandma is saying that Intel made this chip because the A64 is performing so well, etc., etc. That cannot be further from the truth. If this chip is trully going to be demonstrated in a couple of weeks then Intel decided to make this long before AMD had any working silicon of the A64. Long before they could do more than just guess at its performance.

I think you're taking things a bit far in saying that people think that this is a "kneejerk" reaction to AMD's 64-bit performance...

The question being asked is:

(1) Has Intel created their own "x86-64" design from scratch (i.e., the Yahmill project that we have been hearing about for the past 2-3 years) and is now ready to demonstrate it or,

(2) Has Intel been developing a processor using AMD64 extensions during that time instead given the cross-licensing agreement between Intel and AMD

That was the whole point of this thread. I don't think that anyone turned this into a "Fanboy" contest until you mentioned it

 

[DAPUNISHER]

Originally posted by: Alkaline5

Originally posted by: OddTSi
The ignorance of fanboyism really astounds me sometimes.

IMO, the interesting issue here is the fact that Intel has been publically claiming that x86-64 was a waste of time ever since it was announced. Then they turn around and apparently have been working on their own x86-64 processors for years just in case the architecture actually caught on. That's good business strategy (berate the competitor and then cash in on their success at the last second) but they're not going to win any Nobel prizes by talking out of both sides of their mouths.

The public has an extremely short attention span and 99% of the population has no idea Intel ever made those comments so no harm, no foul. I guess some corporate customers might potentially hold it agaisnt them, but I doubt it since they most likely admire the marketing savy and ability to switch gears so fast

 

[dullard]

Originally posted by: Alkaline5
IMO, the interesting issue here is the fact that Intel has been publically claiming that x86-64 was a waste of time ever since it was announced. Then they turn around and apparently have been working on their own x86-64 processors for years just in case the architecture actually caught on. That's good business strategy (berate the competitor and then cash in on their success at the last second) but they're not going to win any Nobel prizes by talking out of both sides of their mouths.

Intel has been saying that the consumer doesn't need 64-bit right now. Intel also said the consumer will need it in the future. Intel then said it will provide a 64-bit processor in the future. How is that "claiming that x86-64 was a waste of time"?

I tend to agree with Intel. The consumer isn't hitting the 32-bit wall yet. Not one person here has even claimed to hit the 4 GB limit, and we are enthusiasts. There is no 64-bit OS that is in wide use, there are no programs that REQUIRE 64-bits either. So now is not the ideal time to have a 64-bit processor release. Give it a year or two and it'll be a completely different story.

Back to the original subject. I was under the impression that Intel and AMD cross license everything. So it would be pointless for Intel to spend the time and money developing a different version of x86-64. They can just slap on AMDs design and everything will work fine. Also I was under the impression that Microsoft said publicly that they will NOT make an OS using a different form of x86-64. Thus even if Intel did attempt to make one, they'd have no popular OS to use.

And how will this help AMD's marketing? They've both shared technology for years...

 

[Hottie]

My guess is it is a last min thing,
1) if Intel have a 64bit down the pipe, they won't said we don't need it until 06(that is 3yrs alway)
2) if Intel have a 64bit down the pipe, MS won't said they will only 1 version of 64bit xp for AMD.

 

[OddTSi]

Originally posted by: NFS4
I think you're taking things a bit far in saying that people think that this is a "kneejerk" reaction to AMD's 64-bit performance...

The question being asked is:

(1) Has Intel created their own "x86-64" design from scratch (i.e., the Yahmill project that we have been hearing about for the past 2-3 years) and is now ready to demonstrate it or,

(2) Has Intel been developing a processor using AMD64 extensions during that time instead given the cross-licensing agreement between Intel and AMD

That was the whole point of this thread. I don't think that anyone turned this into a "Fanboy" contest until you mentioned it

From the Inquirer article that YOU posted to start this thread off:

Intel is having to prepare a 64bit x86 processor. It had no choice; big customers like Dell will have been breathing down its neck for years over the AMD64. It's certainly no coincidence that Dell was dropping big hints about an Intel 64bit x86 processor a short while ago; Dell is feeling the pinch of customers wanting 64bit PC compatible machines and not being able to supply those needs.

That whole paragraph might as well be paraphrased; "this is a kneejerk reaction to the success of the Athlon64". And the reason I replied talking about fanboys is because that is EXACTLY what the Inquirer is, at least when it comes to AMD v. Intel. I think there's maybe a couple of blind, deaf, mute, and illiterate people in some obscure third world country that haven't yet figured out that the Inquirer is the biggest AMD-fanboy site of them all. The rest of us figured it out a long time ago.

 

[OddTSi]

Originally posted by: dullard
Also I was under the impression that Microsoft said publicly that they will NOT make an OS using a different form of x86-64. Thus even if Intel did attempt to make one, they'd have no popular OS to use.

Not exactly. Microsoft said they wouldn't support a THIRD architecture for Intel. The two that they are willing to support are x86-64 and IA64. They kind of hinted that if Intel wanted to make an IA64-based desktop processor they would be willing to make WinXP-64 work with it.

Reprint of Reuter's article regarding Microsoft's announcement

 

[arod]

I guarantee you that if the doom3/HL2 benchies really show a 35% performance increase as rumored/demonstrated AMD64 will be flying off the shelf. If Intel doesnt have something comparable nobody will want to buy an intel machine except those that buy from best buy and the likes but any sertious gamer will buy 64bit after doom3/HL2.

 

[Alkaline5]

Originally posted by: dullard
Intel has been saying that the consumer doesn't need 64-bit right now. Intel also said the consumer will need it in the future. Intel then said it will provide a 64-bit processor in the future. How is that "claiming that x86-64 was a waste of time"?

At the Opteron's release Intel wasn't exactly supporting x86-64 as the next big thing, if you didn't notice. "Wait until we have a 64-bit processor that we say you need" is the message I was getting from their press at around that time. Whether that processor was supposed to be the XeonCT in '04 or something else entirely several years later, I have no idea. Regardless, I doubt anyone could find a quote from Intel praising x86-64 as a breakthrough for the future of computing.

 

[FishTankX]

Originally posted by: arod
I guarantee you that if the doom3/HL2 benchies really show a 35% performance increase as rumored/demonstrated AMD64 will be flying off the shelf. If Intel doesnt have something comparable nobody will want to buy an intel machine except those that buy from best buy and the likes but any sertious gamer will buy 64bit after doom3/HL2.

Most games have about a 35% decrease from running 64 bit..

 

[arod]

Originally posted by: FishTankX

Originally posted by: arod
I guarantee you that if the doom3/HL2 benchies really show a 35% performance increase as rumored/demonstrated AMD64 will be flying off the shelf. If Intel doesnt have something comparable nobody will want to buy an intel machine except those that buy from best buy and the likes but any sertious gamer will buy 64bit after doom3/HL2.

Most games have about a 35% decrease from running 64 bit..

Thats because driver issues havent been ironed out yet, ATI doesnt even have 64 bit drivers out yet nonetheless optimized, at a trade show earlier there were reports of developers saying 35-40% increases in doom3 and HL2 using 64 bit optimizations, dont ask me to look it up because I dont remember where I saw that jsut read it here somewhere, dunno if its accurate or not but interesting still.

EdIt you made me look it up because I wasnt sure...

Will Half-Life 2 have any special optimizations for 64-bit processors?

Gabe Newell: I would expect we would run about 30% faster clock for clock comparing an Athlon running 32-bit code and an Athlon 64 running 64-bit code. Release of the 64-bit client will be gated on MS releasing 64-bit Windows.

http://www.halflife2.net/article_valvespeak_1.php

 

[jm0ris0n]

Well the opteron has been a fully functional chip for like 2 years, correct (since engineering samples were shown) ? This would give Intel plenty of time.

But one has to question why Intel would invent a second 64-bit instruction set completely from scratch. This would be foolish as they already have one. I would like to be optimistic and say they will demonstrate a x86-64 AMD compatible chip. Unless they seriously retooled the itanium, this is their only option.

Wouldn't it be funny if Intel showcased an Itanium/Xeon hybrid and made no mention of the chip coming to the mainstream desktop market because the consumer space 'does not need 64-bit computing'.

Haha

BTW: When is the IDF ?

 

[Accord99]

Originally posted by: arod

Will Half-Life 2 have any special optimizations for 64-bit processors?

Gabe Newell: I would expect we would run about 30% faster clock for clock comparing an Athlon running 32-bit code and an Athlon 64 running 64-bit code. Release of the 64-bit client will be gated on MS releasing 64-bit Windows.

So an A64 of unknown configuration running 64-bit code is faster than an Athlon XP of unknown configuration running 32-bit code by about 30% is no big deal, given that most game benchmarks already give a significant edge to the A64 over AXPs in 32-bit mode. The statement was released last August and September so its not what I call up-to-date.

And with effects-rich games like HL2, people will probably end up being video card limited.

 

[ViRGE]

To answer the original question, Intel is either going to use IA64 or AMD64; for them to introduce a 3rd 64bit design would be insanity on all parts. As to which design they choose, I'm leaning on AMD64, since trying to merge a fast IA64 chip and a fast IA32 chip would result in a chip that's going to be defecient in some noticable way.

 

[RaynorWolfcastle]

NFS4, nobody knows what the tech Intel will be demoing is based on; the best you've got so far are some educated guesses as to where both AMD and Intel are trying to push the market.

IMHO, there are 3 probable possibilities:

A) x86/ IA64 hybrid - I really wouldn't be surprised if Intel pulled this rabbit out of the hat. Considering how long Yamhill has been rumoured to exist, I see this as a definite possibility
B) AMD64 - I'd be surprised if Intel went for a straight implementation of AMD's 64-bit extensions, but it is possible.
C) Intel x86-64 - I think this is much more probable as Intel surely needs some marketing leverage. They make CT technology backwards compatible with AMD64 but add additional instructions for SMT or whatever else they like. This would probably give them marketing leverage and would hurt AMD's campaign as they could use their clout to make the adoption of the other instructions more difficult and make CT appear to be a "better 64-bit" to the average consumer.

If I had to guess, I'd say we'll see either A or C, I just don't see Intel using AMD's instruction set as-is - if only to save face.

 

[NFS4]

Originally posted by: RaynorWolfcastle
NFS4, nobody knows what the tech Intel will be demoing is based on; the best you've got so far are some educated guesses as to where both AMD and Intel are trying to push the market.

IMHO, there are 3 probable possibilities:

A) x86/ IA64 hybrid - I really wouldn't be surprised if Intel pulled this rabbit out of the hat. Considering how long Yamhill has been rumoured to exist, I see this as a definite possibility
B) AMD64 - I'd be surprised if Intel went for a straight implementation of AMD's 64-bit extensions, but it is possible.
C) Intel x86-64 - I think this is much more probable as Intel surely needs some marketing leverage. They make CT technology backwards compatible with AMD64 but add additional instructions for SMT or whatever else they like. This would probably give them marketing leverage and would hurt AMD's campaign as they could use their clout to make the adoption of the other instructions appear to be a "better 64-bit" to the average consumer.

If I had to guess, I'd say we'll see either A or C, I just don't see Intel using AMD's instruction set as-is - if only to save face.

C sounds most probable