Future Athlon 64s to include SSE3 support

[burningrave101]

http://www.anandtech.com/news/shownews.html?i=21618

I swear. What are they going to adopt from Intel next. Why dont they just call them Prescotts lol. I predicted this a few weeks ago in an A64 vs Pentium 4 thread that AMD would resort to useing SSE3 also just like they did MMX, SSE, and SSE2. Everyone tried to tell me SSE3 wasn't important just like SSE2 but even less. But it looks like AMD thinks its important. What i dont get though is why Intel ever signed that cross licensing agreement with them. Its been a one way trade since the beginning. They probably just did it so they could know what AMD is working on and be able to keep a leash on them. Plus if AMD ever went under Intel would become a monopoly and risk losing billions because of our governments laws they would come in and split up the corporation just like they did Microsoft which went from $126 a share to $26 a share all because they became a monopoly and investigations started.

 

[Michaelius]

Hehe I can tell why they aren't called presscotts but that would put me at risk of being banned for starting flame wars

 

[Ugly_Jim]

ia64 comes to mind

 

[MadRat]

AMD has no IA64 license. Plus SSE3 already had 3DNow! compatibility in it, so why wouldn't AMD endorse it? And Intel has a cross licensing agreement with alot of the industry, from big boys like IBM and HP, to the little guys like VIA and Transmeta. Intel has to endorse cross licensing or they'd be in breach of the monopoly laws in the US and EU.

 

[xonik]

The question is, why not adopt SSE3 if you were in AMD's shoes? This isn't an originality contest. If you really want to get into a little comparison, Intel technically bummed the x86-64 extensions off AMD and gets to slide into a compatible OS after AMD did a lot of the initial legwork with MS. Cross-licensing is a great idea because a) it is a good protection against the "Apple effect," that is, making technologies proprietary and incompatible, and b) it saves the involved companies a lot of developmental costs.

And SSE3 still isn't important...right now. No one is going to tell you that it will NEVER be important. That's just foolish.

 

[burningrave101]

Q15: Does Intel's 64-bit extension technology include support for AMD's 3DNOW!* instructions?
A15: No. Intel's IA-32 architecture does not support the 3DNOW! instruction set today. This will not change with the addition of our 64-bit extension technology.

http://support.intel.com/technology/64bitextensions/faq.htm

Well at this point there aren't any facts supporting that Intel adopted AMD's version of x86-64 other then they are alot alike and will be compatible on the same 64-bit OS. Intel has been working on their Yamhill Project which is their x86-64 extensions for years. And as simple as x86-64 extensions are, there's no reason for Intel to just adopt AMD's version. I mean all they are doing is adding 6 more GPR's and 6 more XMM registers and making them 64-bit. So of course they are going to look alot alike. Why wouldn't they.

Intel’s top managers have been considering the enablement of the 64-bit extensions in Prescott and Tejas processors for some time now, but no final decisions have been made. What we know for sure is that Intel is not likely to turn on additional functionality of the Prescott processor until 2005, probably when AMD’s 64-bit processors become more or less wide-spread on the market and may affect Intel’s sales. In early 2005 Intel is set to launch its code-named Tejas CPUs.

According to our information, the extensions may be a part of the well-known Yamhill project and will not be compatible with AMD’s 64-bit extensions available now in AMD Opteron and Athlon 64 processors. One of the first products to sport Intel’s own 64-bit extensions to IA32 may be Intel’s next-generation Xeon MP processor code-named Potomac, another source suggests.

http://www.xbitlabs.com/news/cpu/display/20031216070734.html

Intel joins Advanced Micro Devices in releasing a processor with 64-bit extensions technology, which has attracted the interest of major server vendors such as IBM and Sun Microsystems.

Intel's 64-bit extension technology will be software-compatible with AMD's 64-bit extension technology, Barrett said. A few architectural differences won't affect application compatibility, he added.

The first Intel chip to take advantage of the 64-bit extensions technology will be one code-named Nocona, the next generation of the Xeon DP processors for workstations and low-end servers. Nocona is scheduled for release in the second quarter.

Future versions of both the Prescott processor and the Xeon MP processor will also include this technology. Prescott is the core for existing Pentium 4 CPUs.

Microsoft Windows Server 2003 for 64-bit Extended Systems will support Nocona, said Steve Ballmer, Microsoft chief executive officer, in a videotaped presentation.

Also, Microsoft confirms it is working with both chipmakers to ensure both Intel and AMD CPUs can run the 64-bit extended systems versions of Windows Server 2003 and Windows XP, which will also support both 32-bit and 64-bit applications.

Intel's effort to ensure its processors work with Microsoft's existing 64-bit operating system is good news for customer companies, says Nathan Brookwood, principal analyst at Insight 64.

"Intel will be a uniter rather than a divider, and that's very positive news," Brookwood says.

http://www.pcworld.com/news/article/0,aid,114820,00.asp

All the rumors point to Prescott already having 64-bit extensions built in that can be turned on and off but Intel was not able to enable them because of the fact socket 478 more then likely cannot provide the power needed. A move to LGA 775 would alleviate this issue.

And if you want to get technical all of AMD's processors are using Intel's 16-bit and 32-bit extensions already.

SSE3 will be important once software developers start using it to their advantage and re-write applications to use those extra instructions. And anyone that thinks SSE2 isn't a big deciding factor in the A64's performance is an idiot.

The A64's are essentially K7 cores with on-die controllers and 64-bit extensions which are just added registers. There's nothing revolutionary about them. AMD just decided to go ahead and do what Intel didn't think the desktop consumers were ready for yet. The problem with 64-bit extensions on the desktop is the wide array of hardware devices and the severe lack of driver support for all of it. With servers that issue is far far less extent. And being that Intel is the driving force behind drivers for Windows and always have been its a good thing that Intel decided to announce and release the extensions on their Xeon processors because otherwise there would be very little support to come. ATI couldn't even be bothered to release 64-bit drivers for the AMD processors. That probably has alot to do with Intel and ATI's strong ties togethor. But if Chipzilla moves the others will follow or go bankrupt. AMD doesn't have the fabs or the money to keep up with Intel in a processor war and take any large amount of the market share. AMD has been barely breaking even with their hammer cores. They have been makeing more money off their Athlon XP cores. AMD is only expected to come up with 10,000 socket 939 130 nano chips by the end of third quarter. Far less then what they need. And with 90 nano chips in 4th quarter (supposedly) most people aren't going to buy the 130 nano because they will be more expensive and wont overclock as well. And with all the problems with heat that Intel had from shrinking the die to 90 nano, you can bet AMD with their far more limited budget and research abilities may have those same issues and they might just get delayed till 2005.

 

[xonik]

(this space reserved for later)

 

[FreiDOg]

Originally posted by burningrave101

Well at this point there aren't any facts supporting that Intel adopted AMD's version of x86-64 other then they are alot alike and will be compatible on the same 64-bit OS. Intel has been working on their Yamhill Project which is their x86-64 extensions for years. And as simple as x86-64 extensions are, there's no reason for Intel to just adopt AMD's version. I mean all they are doing is adding 6 more GPR's and 6 more XMM registers and making them 64-bit. So of course they are going to look alot alike. Why wouldn't they.

Don't give me that marketing bullshit.

To quote Linus Torvalds, who you know actually might know something about this:

What I found so irritating is that _hours_ after the Intel announcement, people were _still_ confused about whether the new intel chip was actually compatible with AMD's chips. Why the f*ck not just come out and say so, and talk about it? It took people actually reading the manuals (which didn't mention it either) to convince some people on the architecture newsgroups that yes, 'ia32e' was really the same as 'amd64' except in the small details that have always set Intel and AMD apart.

IA32 is x86-64 (from any relevant software perspective), and Intel won't even mention AMDs part in it.

The conversation was posted on KernelTrap.

SSE3 will be important once software developers start using it to their advantage and re-write applications to use those extra instructions. And anytone that thinks SSE2 isn't a big deciding factor in the A64's performance is an idiot.

SIMD, any decent SIMD, is a big deal, in the 1% of so of all software that uses it.
SSE3 may be more popular because it includes some very nice functions for 3D, like vector math on well, vectors dot products, cross products ect.
AMD would be plain and simple stupid not to take advantage of what they are (essentailly) paying for through the cross liscenses of propriatary extenstions.

though i guess to use the phrase of the week, right now it's just blue crystals.

 

[burningrave101]

IA32 is x86-64 (from any relevant software perspective), and Intel won't even mention AMDs part in it.

Well to be technical, IA-32 itself is just an x86 32-bit processor. IA-32e which is the 32-bit processor with 64-bit extensions is x86-64.

It took people actually reading the manuals (which didn't mention it either) to convince some people on the architecture newsgroups that yes, 'ia32e' was really the same as 'amd64' except in the small details that have always set Intel and AMD apart.

Even if their alot alike except for small details why does that mean AMD had any part in it. 64-bit extensions are 64-bit extensions and if your going to make them compatible on the same OS they can't be totally different. Nobody seems to think its a big deal that the A64 is using a copy of Intel's 32-bit extensions.

Coodos to AMD if they did play a role in Intel's 64-bit extensions but it just goes against Intel's history and the way they do things.

 

[0ldman]

If AMD did it first, AMD created x86-64, and now Intel uses it, how is Intel NOT copying AMD?

It isn't a big deal, AMD made it first, Intel had to choose, create something entirely different or use an existing ISA.

If Intel does it better, which wouldn't surprise me, then yea, its a big deal.

 

[Merlin45]

yes there is a lot of copying going on, and it is all perfectly legal, x86-64 is one of the very rare instances where it is intel copying amd and not the other way around though. in most cases, (mmx, sse, sse2, and now 3, not to mention the x86 architecture itself) it has been amd using intel's stuff.

 

[Mark Larson]

I'll stay out of this thread.

 

[Josh_B]

Originally posted by Merlin45
yes there is a lot of copying going on, and it is all perfectly legal, x86-64 is one of the very rare instances where it is intel copying amd and not the other way around though. in most cases, (mmx, sse, sse2, and now 3, not to mention the x86 architecture itself) it has been amd using intel's stuff.

Well Intel is also using Alpha technology in the Itanium.... whats the big deal?

Everyone copies from each other...

 

[burningrave101]

Originally posted by Mark Larson
I'll stay out of this thread.

Good idea

 

[FreiDOg]

Originally posted by burningrave101
Well to be technical, IA-32 itself is just an x86 32-bit processor. IA-32e which is the 32-bit processor with 64-bit extensions is x86-64.

and?

Even if their alot alike except for small details why does that mean AMD had any part in it. 64-bit extensions are 64-bit extensions and if your going to make them compatible on the same OS they can't be totally different.
Coodos to AMD if they did play a role in Intel's 64-bit extensions but it just goes against Intel's history and the way they do things.

Of course they used AMDs specifications (published openly in Oct 2000, long before Yamhill or *T or IA32e was ever even heard of), if they didn't use AMD's specifications they wouldn't be compatible with AMD's solution.
2 ISA extensions as large and complex as AMD64 / IA32e just don't happen to be completely intercompatible.

Nobody seems to think its a big deal that the A64 is using a copy of Intel's 32-bit extensions.

you obviously do, or you wouldn't have made a thread crtizising AMD for adopting SSE3....


And for the record, i don't think it's a big deal intel is using x86-64, it's a natural evolution, it's a nice marketing point, and IA64 is really unsuited to replace x86 right now,
but i consider it very petty that they refuse to acknoledge AMD's role in it's development.

 

[burningrave101]

Originally posted by FreiDOg
you obviously do, or you wouldn't have made a thread crtizising AMD for adopting SSE3....

My dispute is more with the way the media is portraying all of this out to look like more then i care what AMD is adopting. They brush off everything AMD is useing from Intel and point out one thing Intel MIGHT be useing off of AMD and it just sturrs the fire even more for all the flaming AMD idiots out there that already perceived AMD as being holier then holy and so far beyond Intel because of the fact their processors are more "efficient" when they dont even realize that a processor with a slower clock speed thats more efficient isn't any better then a processor with a much faster clock speed and less efficient. And they shouldn't even refer to it as efficiency because it all has to do with how much work gets done on each pipeline stage. The AMD processor can get the work done very quickly at slower speeds but has to do alot more work per stage while the Intel processor can do alot less work per stage and move alot more data through if the speed is fast enough. But you also have to take branch prediction and all that into consideration because of the longer pipeline stages it also takes longer to clear all the data out and start over if there is a mistake made but with Prescotts improved brand prediction there should be alot less issue with that.

Anyways, back to what i was talking about lol.

Back before the Hammer core was released, the AMD fanboys out there were always saying how AMD wouldn't be petty enough to need SSE2 and then came along the Hammer core which would include the SSE2 instruction set. Then they said that SSE3 was pointless and was only 13 new instructions and was nothing more then an addon to the SSE2 instruction set and that AMD doesn't need it. Well now AMD is adding this also into their core. Intel could add 3DNow! into their Prescott core but they didn't. If Intel was benefiting more from the technology AMD was releasing then i could understand all this alot better. You can say Intel is getting x86-64 from AMD and even if they did base their exensions off of it, it was only for compatibility reasons so that it would run fluently on the same OS. Intel did 16-bit and 32-bit extensions not to mention the entire x86 architecture. Intel could do their own 64-bit extensions in a heartbeat. I'm not blaming AMD for doing this because they have the cross-licensing agreement with Intel so they are entitled to use it if they choose. I just have a hard time putting up with alot of this crap thats floating around on the net about Intel bowing down to AMD now and adopting x86-64 as if it was anything near everything AMD has kissed Intel's ass and used lol. You dont see AMD giving Intel recognition for anything their using in their core that came from Intel devlopment. If you went up and asked a nOOb that loved AMD about the instruction sets and the x86 architecture or whatever else he would just presume AMD had done it all themselves. So why should Intel give AMD any recognition for doing one thing right that was worth adopting if they even completely did that.

What alot of AMD fanboys dont realize is that alot of this just boils down to the fact Intel has to have AMD around because if all of a sudden AMD went under, Intel would be hit from a thousand different directions from lawsuits over everything imaginable and would end up costing Intel billions of dollars. Its not illegal to become a monopoly but if you do anything to force others into buying your chip because you own the majority of the market over a smaller dealer then your open for a serious lawsuit. And if AMD all of a sudden wasn't here, everyone would think that is what had happened. Its happened to Microsoft and alot of other large companys.

And they also dont realize that AMD would never be able to push 64-bit mainstream by themselves. AMD brought 64-bit to the desktop first but Intel will take it to the multitude as a whole and will provide the driver support thats neccessary. Software developers will not base their work on what AMD can provide because AMD has yet to prove they can provide enough chips in any large quantity and do it on time.

Personally i think the A64's are great processors and that AMD needs to stay around if for no other reason than to just keep Intel's prices down. I just get tired of people overlooking the facts and only seeing what they want to see.

And judgeing from some of your responses there must be about as many AMD lovers lurking in the Intel section as the AMD section lol.

 

[FreiDOg]

fanboys are idiots by deffinition. Don't get so rilled up when they prove it.

 

[2phastPRO]

Seems to be a whole lotta nothing and fanboys here.

To address a much less complex note; its quite obvious that AMD has forced intels hand into the 64bit round concerning home usage. Lets try and give the little guy some credit as they are doing a superb job at making Intel look like a bunch of monkey's stuck out on a boat out in the middle of the Marina Trench waving signs with HyperThreading marked all over them.

GIVE them their due credit as they have THE fastest chips out on the market as of now. (I dont want to see any replies with "Wheres your proof?" Its all over the god damn net )

DO NOT give the ussual bullshit; "Who needs 64bit now" or "Gee look they're adopting MORE stuff from Intel" Thats a bit "fanboyish" and a bit over protective of a company and shows people just how hurt you are that the opposition has one two'd your "precious" party.

Thank You for your time.

 

[Merlin45]

I just want to point one thing out, iirc 3dnow is MMX in everything but name, and 3dnow pro is SSE, so in a way, intel already has both.

 

[burningrave101]

Originally posted by 2phastPRO
GIVE them their due credit as they have THE fastest chips out on the market as of now. (I dont want to see any replies with "Wheres your proof?" Its all over the god damn net )

DO NOT give the ussual bullshit; "Who needs 64bit now" or "Gee look they're adopting MORE stuff from Intel" Thats a bit "fanboyish" and a bit over protective of a company and shows people just how hurt you are that the opposition has one two'd your "precious" party.

Thank You for your time.

Well i'll give the A64 credit that its the best affordable gaming processor and its better at some other things like office apps but the Pentium 4 Northwoods are still better overall. They win in 60% or more of the benchmarks and lead in encodeing, rendering, multitasking, applications that are multithreaded, and overclocking. Not to mention Intel chipsets are rock solid and dependable while the A64 chipsets are far from it. And then we could get into all the memory incompatibility on A64 boards thats rampant. You can't base the best processor on just gaming performance especially when in most games the difference is by very little and the Pentium 4 wins in some games. And all prices aside the Pentium 4 3.4EE is the fastest processor out even in gaming.

The performance between a Pentium 4 3.0c and a A64 3000+ is roughly 60/40 to 65/35 judgeing from an large overview of like 20 different review scores.

A 3.0c @ 3.6Ghz will beat a 3000+ @ 2.2Ghz easily overall and in alot of games too. 200Mhz is about all you can get out of an A64 on average on normal cooling. Some get a little more. And some get a little less.

Also AMD is having to ramp up their processors clock speed almost on par with the increase in clock speed of an Intel processor. So they aren't nearly as efficient as the Athlon XP processors which only saw less then 100Mhz increases per 200Mhz increase in Intel rated performance. Of course AMD's whole rating system is pretty wacked out too.

A64 3000+ 2.0Ghz
A64 3200+ 2.0Ghz
A64 3400+ 2.2Ghz
A64 3700+ 2.4Ghz

I can show you plenty of links if you dont want to just take my word for it though.

Heres a good one:

http://www.techreport.com/reviews/2004q1/athlon64-3000/index.x?pg=1

So whats sadder. The fact that Intel is getting driven into producing x86-64 CPU's or the fact that AMD A64's are slower then older Northwood cores overall.

But it really all comes down to what you use the PC for. If your just a hardcore gamer then the A64 is best and if your looking for something thats rock stable and performs best overall then its the Pentium 4.

 

[Grazehell]

Originally posted by burningrave101
http://www.anandtech.com/news/shownews.html?i=21618

I swear. What are they going to adopt from Intel next. Why dont they just call them Prescotts lol. I predicted this a few weeks ago in an A64 vs Pentium 4 thread that AMD would resort to useing SSE3 also just like they did MMX, SSE, and SSE2. Everyone tried to tell me SSE3 wasn't important just like SSE2 but even less. But it looks like AMD thinks its important. What i dont get though is why Intel ever signed that cross licensing agreement with them. Its been a one way trade since the beginning. They probably just did it so they could know what AMD is working on and be able to keep a leash on them. Plus if AMD ever went under Intel would become a monopoly and risk losing billions because of our governments laws they would come in and split up the corporation just like they did Microsoft which went from $126 a share to $26 a share all because they became a monopoly and investigations started.

sorry but when I read this excuse of a post it just get me irrationally angry.
Maybe AMD and Intel exchanged Technologies under their cross licensing. SSE3 for x86-64...wow you think that could be possible?

 

[burningrave101]

Originally posted by Grazehell
sorry but when I read this excuse of a post it just get me irrationally angry.
Maybe AMD and Intel exchanged Technologies under their cross licensing. SSE3 for x86-64...wow you think that could be possible?

Sure its possible but probably not. What did Intel get out of everything else AMD is useing? And even if they didn't have a cross licensing agreement, Intel still could of done 64-bit extensions that are compatible on the same OS as the A64's. If you'll look on Intel's site and other sites that reported on the Intel's newly anncouned 64-bit extentions they refer to them being alot alike with slight differences. Intel did the entire x86 architecture for christ sakes. They own the rights to all of it. Adding a few extra registers to take it from 32-bit to 64-bit is nothing.

I just get tired of Intel coming out with something and then all of a sudden the AMD processors have the exact same thing implemented into their core and then AMD fanboys bash all over Intel. Like i said, im mad at the ignorant media and fanboys far more then i am at AMD for useing it.

And why do you guys care so much about AMD anyways. Have you forgotten what section your in?

 

[Mark Larson]

Originally posted by burningrave101
And why do you guys care so much about AMD anyways. Have you forgotten what section your in?

Oops i forgot i was in the fanboy circle-jerk section.

 

[Merlin45]

the only thing I find interesting about this is that AMD had the decency to cite SSE3's proper name, Prescott New Instructions, where as Intel went to great lengths to avoid connecting IA-32e with AMD, despite the fact that it is a slightly changed version of x86-64

 

[burningrave101]

Originally posted by Mark Larson
Oops i forgot i was in the fanboy circle-jerk section.

Sorry your in the Strictly Intel section. That other section is one click down.

But i like to refer to it as:

"The only section where fanboys will start new threads just to flame you".

The whole reason i posted this in the Strictly Intel section instead of General Hardware is cause i wanted to minimize the AMD fanboys and express how i felt about this and see if anyone else had similar feelings or ideas on the subject matter.

 

[Mark Larson]

So you only want to hear what you already believe, and will put your fingers in your ears and scream "la la la la" when someone points out that you're wrong.

 

[burningrave101]

Originally posted by Mark Larson
So you only want to hear what you already believe, and will put your fingers in your ears and scream "la la la la" when someone points out that you're wrong.

No see thats where your wrong. I'll listen to anything if you can prove it. But like you most AMD fanboys talk alot and prove very little. If you go and look at any of the A64 vs Pentium 4 threads i've participated in i've provided ALOT of links and information to back up what im saying. I dont try to force someone to believe things that can't be proven and aren't true. So where exactly have you pointed out that im wrong BTW? Or anyone else in this thread. Everyone is just stateing their opinions on the matter also and some of them obviously care for AMD over Intel even though their in the Strictly Intel section. I never go into the AMD section trying to talk people into believing something about Intel processors unless that thread starts somewhere else and gets moved there like what happened a while back but i dont even follow them in there any more because alot of you are extremely childish and refuse to hear the truth about anything. I dont care if someone proves me wrong if they can actually prove it with information thats real instead of what they think. I want to learn. I dont want to base my knowledge on opinions like so many forums and review sites out there who fall into that category. I have nothing against AMD, i just dont agree with everything thats going on. And i definitely dont have a single thing against AMD users cause their great processors. I do have a thing against AMD fanboys though that are total idiots to the point they refuse to listen to anything even if there are 500 links showing the exact opposite of what they want to believe.

It was pretty funny in the AMD section a while back because that one guy got to take a vacation for a while because he kept flaming me lol.

So if you want to try and prove me completely wrong about things then i hope you do, but if your just going to go on and on and on and on about how you think im a fanboy why dont you just keep it to yourself cause your not proving a thing.

 

[0ldman]

Frankly I can't think of a post where I read that AMD was too good to use SSE2... name one here.

Not AMDzone, or some other site, right here on Hardforum.

Secondly, in this business, if ANYONE has a one up on you, if you do not reply accordingly you will soon be the underdog. AMD's 64bit is gaining momentum. Intel has to answer this. Same as MMX, AMD had their version.

The absolute *only* reason this seems to be a big deal is that for the first time I can remember Intel is using AMD's tech rather than the reverse. AMD was the inovator in this instance, Intel was left to reply.

Kudos to AMD, nice job. Don't get lazy or Intel will bite you in the ass.

Where is there a problem here?
Technology moves forward. If AMD couldn't get the support needed to push x86-64, why is Intel going forward with it given their previous comments toward 64bit on the desktop? Remember, we don't need it, no reason for it.

AMD got something right for a change, Intel's hand was forced and it got your panties in a bunch, otherwise you wouldn't have started this thread. Lose the brand loyalty, blinders don't help anyone.

I can't say if this was for the best, x86 ISA has its issues, it seems that keeping our backwards compatability has its costs. We can sacrifice old programs and have a new ISA, IA64 Intel was originally pushing, or x86 can be extended and chip designers cuss and software writers can deal with it while we continue to trod along the evolution of x86, old and new, good and bad.

 

[2phastPRO]

I totally agree with you on that note 0ldman

 

[Josh_B]

Originally posted by burningrave101
No see thats where your wrong.

Actually, YOU'RE (as in 'you are') wrong.

Intel and AMD cross license each other... so who cares? Are you going to criticize Transmeta and VIA for using IA32 as well?

 

[burningrave101]

Originally posted by Josh_B
Actually, YOU'RE (as in 'you are') wrong.

Intel and AMD cross license each other... so who cares? Are you going to criticize Transmeta and VIA for using IA32 as well?

No, because i havn't run into many Transmeta and VIA fanboys lately lol.

I'm just fed up with AMD fanboys that think AMD is everything and Intel is scum. And believe me, alot of them do think that. I was just in another hardware channel today where there were a handful of AMD freaks that were mouthing off about how much better AMD was than Intel and that they would never own an Intel processor. They seem to forget a majority of that processor their using is based off of Intel technology. And after seeing the news posted about AMD adding SSE3 to their processors when they just got done adding SSE2, i couldn't help but say something because of all the crap thats going on about x86-64 extensions on the net.

 

[leukotriene]

And believe me, alot of them do think that.

I'm not sure why anyone should believe you on anything at this point, especially given your first post in this thread. Everyone (excuse me, everyone not on crack) knows about the cross-licensing deal that has existed for many years, and everyone and their sister's dog knows that AMD using SSE3 will only make SSE3 adoption faster.
SSE3 adoption being accelerated sounds like a good thing for the consumer to me, but maybe I'm too much of a fanboy

IA-32e is for all intents and purposes a direct copy of x86-64, according to anyone with any knowledge of the specifics involved.
You are not one of those people, nor do I think, given your claims in previous discussions, that you will ever be.

The most humorous part about IA-32e for me is how often and how recently Intel claimed it just "wasn't necessary".

This doesn't mean Intel is a bad company, or you that are a doo-doo head for being such an outspoken Intel devotee (to the point of error, repeatedly), but it does mean Intel is not omniscient and still responds to market demand, which is healthy for a company in a capitalist economy.

In this case the demand was created in large part by its sole credible competitor, but Intel is ashamed and afraid to admit it because admitting it would instantly lend credibility to the competition.
I personally don't think they should worry about it because it won't do anything to their marketshare, but I suppose the more devoted members of their fanbase would consider it akin to being kicked in the testicles.

By the way, you do realize IA-32e doesn't have 64 bits of physical memory adressing?
"It's not REALLY 64 bit then!"
ROFLMAO!
How quickly we forget.....

For anyone who is honestly interested in the notable differences between IA-32e and AMD64, the main page of aceshardware has the very short list.
The primary differences appear to be exceptionally minimal. Which is good for the consumer, but very very very bad for anyone who tries to claim (ahem!) that Intel has come up with a wholly unique solution.

 

[0ldman]

Originally posted by burningrave101
No, because i havn't run into many Transmeta and VIA fanboys lately lol.

I'm just fed up with AMD fanboys that think AMD is everything and Intel is scum. And believe me, alot of them do think that. I was just in another hardware channel today where there were a handful of AMD freaks that were mouthing off about how much better AMD was than Intel and that they would never own an Intel processor. They seem to forget a majority of that processor their using is based off of Intel technology. And after seeing the news posted about AMD adding SSE3 to their processors when they just got done adding SSE2, i couldn't help but say something because of all the crap thats going on about x86-64 extensions on the net.

Take your fanboy rantings to a forum where someone cares.

I don't care what you have heard fanboys spout. For one, who cares? Chevy vs Ford... AMD vs Intel, ATI vs Nvidia... it never ends.
Secondly, you are hardly in the position to rant about fanboyism. Your answer to AMD fanboys is to start threads baiting fanboys.

Go to store.
Purchase clue.
Get life.

Good day.

 

[burningrave101]

Originally posted by leukotriene
By the way, you do realize IA-32e doesn't have 64 bits of physical memory adressing?
"It's not REALLY 64 bit then!"
ROFLMAO!
How quickly we forget.....

Your right its not "TRUE" 64-bit . 64-bit Math, 64-bit Chunks, but not 64-bit Addressing. I dont think anyone ever tried to say that it was.

IA-32e is for all intents and purposes a direct copy of x86-64, according to anyone with any knowledge of the specifics involved.
You are not one of those people, nor do I think, given your claims in previous discussions, that you will ever be.

Yea its pretty hard for anyone to prove facts to an AMD fanboy with just supporting links alone. It usually takes a small act of God .

For anyone who is honestly interested in the notable differences between IA-32e and AMD64, the main page of aceshardware has the very short list.
The primary differences appear to be exceptionally minimal. Which is good for the consumer, but very very very bad for anyone who tries to claim (ahem!) that Intel has come up with a wholly unique solution.

I never said they were completely different or that Intel didn't do their 64-bit extensions nearly identical to the A64's extensions for compatibility reasons. But there are slight differences. AMD does deserve credit for being the backbone of the x86-64 architecture and pushing Intel into making a decision on it though.

And yes everyone knows Intel and AMD have a cross-licensing agreement but the fact remains that AMD has brought very little to the table that Intel needed up till now.

And dont even start on that gay thread in the AMD section lol. There were idiots in there trying to argue with me that the A64 didn't use extensions and that it was "full" 64-bit lmao. And if i remember right, it wasn't me that got to take a forum vacation for improper forum conduct but one of your fellow AMD lovers lol.

Only mistake i made was with not doing more research on the 64-bit extensions AMD did use before hands so i quoted a false fact about them being able to only work 40-bit chunks instead of 64-bit.

Your answer to AMD fanboys is to start threads baiting fanboys.

Yea and it looks like I managed to lure a whole slew of them into this thread now didnt I lol. I thought the mods made it pretty clear a while back that they wanted the opinionated fanboys to stick to their sections and not cross over for disbutes. You havn't seen me wasting my time in the AMD section trying to argue with the rest of you over Intel vs AMD since then. Maybe some of you should take that same advice.

 

[Koz]

Originally posted by burningrave101

And believe me, alot of them do think that.

A LOT... TWO WORDS FOR THE LOVE OF GOD. It really bothers me when people think its one word .

Originally posted by burningrave101

They seem to forget a majority of that processor their using is based off of Intel technology. And after seeing the news posted about AMD adding SSE3 to their processors when they just got done adding SSE2, i couldn't help but say something because of all the crap thats going on about x86-64 extensions on the net.

What you seem to be missing here is the way in which they were implemented. AMD says their processors support SSE, SSE2, SSE3, MMX and x86-64 (AMD64 Extensions). Intel however, uses IA32e to describe what boils down to AMD64 Extensions. They give NO credit where credit is due whereas AMD consistently cites Intel as having given the technology. They refuse to admit AMD has any effect on their plan as a company despite evidence such as their consistent rejection of x86-64 until AMD won numerous design wins from companies such as Sun, IBM, HP, Cray etc. And as to AMD using Intel's technology: They have the agreement to share the technology and Intel puts tons of money in to get software companies to adopt their special instructions (MMX, SSE, SSE2) in their software. AMD would be stupid NOT to add them, and if Intel clears the path for them for free performance increases, why not take the gift. As for Intel not using 3DNow! and 3DNow! Pro (or + whatever its called), MMX and SSE were too similar to justify the inclusion. Intel's $$ and larger market share however made MMX and SEE prevailant. AMD adopted these to acomplish what they had intended 3DNow! to do but they couldnt drop 3DNow altogether because of the backwards compatability; AMD needed to be consistent.

You are completely an Intel fanboy and its to the point where you abandon rationality to support your cause. You complain about AMD fanboys making ridiculous claims, but you yourself are just as bad an Intel fanboy. I don't see how people can get so fervently behind a company. You buy from them to meet your needs, and whichever acomplishes that most efficiently and most cost effectively (or just has the best raw performance if you have the cash ) should be your choice. You posted this thread with the obvious purpose of causing conflict (or so it appears) and so you've acomplished what you wanted. You take pride in getting someone banned which is the wrong approach to take in the forums. The forums are a community and need to be treated as such. Inciting conflict is bad for everybody....

Now back to the topic?...

AMD is slightly ahead of Intel in this current generation of processors. You can't compare overclocked processors because overclocks aren't guaranteed performance jumps. You COULD hit 3.6 GHz on your processor, but at the same time, you could only get to 3.2. Both are possible and both happen, so you cant take the performance jump for granted. In the same light, a AMD processors COULD get a 200MHz overclock, it could get a 400MHz overclock. It all depends on your chip.

Right now the more all around balanced chip IMO is AMD's with Intel's strong point being Encoding/Streaming, which is exactly what its Netburst architecture was designed to do. AMD's strengths are in office productivity and gaming. Intel seems to consistently prevail in synthetic benchmarks, which personally I think should be closely scrutinized, as some of them do not actually provide an accurate picture (PCMark2004 is a perfect example).

 

[burningrave101]

Originally posted by Koz
You are completely an Intel fanboy and its to the point where you abandon rationality to support your cause. You complain about AMD fanboys making ridiculous claims, but you yourself are just as bad an Intel fanboy. I don't see how people can get so fervently behind a company. You buy from them to meet your needs, and whichever acomplishes that most efficiently and most cost effectively (or just has the best raw performance if you have the cash ) should be your choice. You posted this thread with the obvious purpose of causing conflict (or so it appears) and so you've acomplished what you wanted. You take pride in getting someone banned which is the wrong approach to take in the forums. The forums are a community and need to be treated as such. Inciting conflict is bad for everybody....

And whats so rediculous about my claim? I'm not claiming things that aren't true and fanboys usually resort to claiming false facts and opinions about one processor or the other. I favor one processor over the other yes. Why? Because it does its job better then the other processor IMO and thats my opinion and your free to yours. I'm not trying to get anyone banned or start flame wars. I only wanted fellow Intel users that might possibly feel the same way to respond. I wasn't looking for all the AMD lovers to come out of their section and think they need to prove something.

Whenever someone says they want the ultimate gaming machine and thats the main thing they will ever use it for then i tell them to get an A64 or P4EE depending on if they want an affordable chip or if they have money to burn. And whenever they ask for a low budget chip with great performance i say the Athlon XP. Especially the Athlon XP-M right now. Planning on picking one up myself. There's nothing fanboyish about what im saying. It just gets enterpreted that way because im very passionate in the way that i believe and i get very irritated with ignorant people that will support things without seeing the truth behind it all. I agree i should of just kept my thoughts about AMD using the SSE3 instruction set to myself and not get all hot about it but in the same way all these AMD followers that have posted in this thread should of kept to their section unless they were here in favor of Intel because this section is the STRICTLY INTEL section. Its the whole reason i posted in this section. If the purpose of this thread was to get the attention of AMD users i think i would of posted in the General Hardware section or hell even the AMD section lol.

AMD is slightly ahead of Intel in this current generation of processors. You can't compare overclocked processors because overclocks aren't guaranteed performance jumps. You COULD hit 3.6 GHz on your processor, but at the same time, you could only get to 3.2. Both are possible and both happen, so you cant take the performance jump for granted. In the same light, a AMD processors COULD get a 200MHz overclock, it could get a 400MHz overclock. It all depends on your chip.

Right now the more all around balanced chip IMO is AMD's with Intel's strong point being Encoding/Streaming, which is exactly what its Netburst architecture was designed to do. AMD's strengths are in office productivity and gaming. Intel seems to consistently prevail in synthetic benchmarks, which personally I think should be closely scrutinized, as some of them do not actually provide an accurate picture (PCMark2004 is a perfect example).

The A64 is a strong chip. Its the best gaming processor and strong in office applications. But despite all that the Pentium 4 is still a better overall processor. It wins in more benchmarks. And the gaming performance difference is very negligible. What do you think your going to notice more? A couple more frame rates per second or quicker load times in programs like Photoshop and encodeing times when your encodeing a DVD? Synthetic benchmarks asside, most applications are optimized for the Pentium 4 and run best on it because of the Hyperthreading and greater bandwidth. I never figured Overclocking into the performance ratio of the processors. The 3.0c vs the A64 is still 60/40 or more based on more then 20 reviews. If you dont believe me then go dig up that long ass thread in the AMD section where i posted all those tons of links and still noone wanted to believe anything because truth and facts aren't enough. It all boils down to what you primarily use the PC for but for best performance overall today its the Pentium 4 not from my opinion but judgeing from the slew of performance tests and results. If for no other reason the Intel motherboards are very solid and dependable. The A64 chipsets are very poor IMO.

 

[Koz]

Here lets look at Hard|OCP's own most recent A64 benchmarks:
MP3 Encoding:
http://www.hardocp.com/image.html?image=MTA3NzM5MDYwNVllc3gwbWFnTzFfNV80X2wuZ2lm
Aha! A 5 second difference, there's the stuff you notice.
BTW, thats comparing the 3.4EE to a 3400+, which is $600 cheaper, a fairer comparison would be to the FX-53, which is supposed to come out this month. I'd compare it to the 3.4E as its Intel's next gen comparable processor for the average joe, but unfortunately its not with the other benches .

Video Encoding:
http://www.hardocp.com/image.html?image=MTA3NzM5MDYwNVllc3gwbWFnTzFfNV81X2wuZ2lm
Hmm, AMD wins by ~5 minutes. Thats a noticeable difference.

Video Encoding 2:
http://www.hardocp.com/image.html?image=MTA3NzM5MDYwNVllc3gwbWFnTzFfNV81X2wuZ2lm
Intel wins by 6.6 seconds, or less if compared to the 3.4EE.

Now to gaming benches,
Serious Sam and JK2 are high FPS for whatever system you use, so they dont really show either processor making a difference, although AMD wins both, even with a 70FPS gap, its meaningless because you probably wont see the difference between 190 and 260...

Comanche 4:
http://www.hardocp.com/image.html?image=MTA3NzM5MDYwNVllc3gwbWFnTzFfNl80X2wuZ2lm
Pretty balanced as the 3.2's fall behind the 3400+'s which isnt really a fair comparison but I doubt a 200MHz increase would result in a 20 FPS increase for the processor. from 56 to 73 is highly visible and any average gamer could see this difference.

UT2K3:
http://www.hardocp.com/image.html?image=MTA3NzM5MDYwNVllc3gwbWFnTzFfNl81X2wuZ2lm
Another ~20 FPS difference... Most people could probably tell the difference here; I know I could at least.

The synthetics go both ways, but I dont even consider them as in the end, you dont buy a system to run prepackaged attempts to simulate real performance.

 

[burningrave101]

UT2K3:
http://www.hardocp.com/image.html? ... /> l81X2wuZ2lm
Another ~20 FPS difference... Most people could probably tell the difference here; I know I could at least.

Gaming Performance:
UT2K3: I do hope you weren't comparing the flyby performance lol. You seriously think you can tell the difference in 20 fps when the fps is over 200? Its been proven before the human eye can only differentiate around 70-80 fps usually at most. Once it starts getting over 100 fps you might as well call it quits.

http://techreport.com/reviews/2004q1/athlon64-3000/index.x?pg=5

Pentium 4 3.2Ghz wins against the 3200+ in Wolfenstein and Comanche 4. It also beats the 3000+ in Quake III. UT2k3 botmatch the 3200+ is only ahead by 6 fps.

3d max rendering:
http://techreport.com/reviews/2004q1/athlon64-3000/index.x?pg=8
A64 wins one and the Pentium 4 wins one.

Lightwave rendering:
http://techreport.com/reviews/2004q1/athlon64-3000/index.x?pg=9
Pentium 4 3.2Ghz wins all three.

POV-Ray rendering:
Not multithreaded so the A64 tends to do better.

Cinebench 2003 rendering and shading:
3.2Ghz wins 2 and A64 wins 2 in OpenGL shading.

Encodeing:
http://techreport.com/reviews/2004q1/athlon64-3000/index.x?pg=7
3.2Ghz wins in both Lame MP3 and DivX.

More gaming performance:
http://www.hexus.net/content/reviews/review.php?dXJsX3Jldmlld19JRD03MjImdXJsX3BhZ2U9Nw==
3.0Ghz wins in Comanche 4 and Quake III. Trails the 3000+ by 2.5 fps in Call of Duty and 5 fps in X2 the Threat.

WAV to MP3 Encodeing:
3.0Ghz beats 3000+

Xvid Encodeing:
http://www.hexus.net/content/reviews/review.php?dXJsX3Jldmlld19JRD03MjImdXJsX3BhZ2U9Ng==
3.0Ghz wins against the 3000+

Dude i've taken all of these reviews before and compiled them and gotten an average and its around 60% or more in favor of the Pentium 4 in benchmarks and performance tests.

Look at all these reviews i've posted. I'm not just makeing up this crap off the top of my head. I read ALOT of reviews on a daily basis. I dont go off of what one or two hardware sites post about something.

http://www.hardforum.com/showthread.php?s=&threadid=720413&perpage=20&pagenumber=5

And then after you get done with that take a look at how sad some of these AMD users are in that some of them have to resort to personal attacks against me even to the point of new threads calling me out trying to get me to argue more all because i was providing facts and information that they didn't like. I havn't tried to attack anyone in this thread or other threads.

http://www.hardforum.com/forumdispl...=&sortfield=lastpost&perpage=40&pagenumber=11

I didn't start this thread to get AMD users opinions and thoughs. If i wanted them i would of posted in the appropriate place.

http://www.hardtecs4u.com/reviews/2...3400/index4.php

Pentium 4 3.2Ghz wins 9 times. A64 3200+ wins 8 times.

http://www.sudhian.com/showdocs.cfm?aid=478

Pentium 4 3.2Ghz wins 20 times. A64 3200+ wins 13 times. 1 tie.

*I excluded the thorough Unreal Tournament 2003 bench cause it was so extensive. You'll have to just look over it. I'm sure the A64 was ahead through most of it.

http://www.xbitlabs.com/articles/cp...lon64-3400.html

3.2Ghz P4 wins 18 times. A64 3200+ wins 21 times. 1 tie.

http://www6.tomshardware.com/cpu/20040106/index.html

3.2Ghz Pentium 4 wins 32 times. A64 3200+ wins 9 times.

http://www.techreport.com/reviews/2...00/index.x?pg=1

The Pentium 4 3.2Ghz wins 31 times. The A64 3200+ wins 19 times.

http://www.hothardware.com/hh_files/CCAM/amd_a64_3400+(2).shtml

The Pentium 4 3.2Ghz wins 18 times and the A64 3200+ wins 3 times.

http://www.firingsquad.com/hardware...64_3400_review/

The Pentium 4 3.2Ghz wins 18 times. The A64 3200+ wins 10 times.

http://www.amdmb.com/article-display.php?ArticleID=277

Pentium 4 3.2Ghz wins 20 times. A64 3200+ wins 24 times.

http://www.extremetech.com/article2...,1426423,00.asp

Pentium 4 3.2Ghz wins 12 times. A64 3200+ wins 9 times. 1 tie.


No two reviews are ever going to be the same. Peformance cannot be based off one review or one benchmark. They are both good processors and both are better at different things. If your a gamer the A64 is the best choice. If your into encodeing, applications, music, CAD and Photoshop and so forth then the Pentium 4 with Hyperthreading is the best choice. Overall i say the Pentium 4 is the best cause it usually wins in more benchmarks, its alot more overclockable, its cheaper to buy a 2.8c and overclock then a 3200+, and main reason i like it is that Pentium 4's work really well at multitasking. They also perform nearly as well in most gaming benchmarks where the A64 wins by just a few points.

And you can try to exclude overclockability all you want because of its variability but high overclocks have become the common thing now on Pentium 4's whereas on the A64's is anything but. 200Mhz is average. Thats it. Sure some people get more but the average is all the matters. The average for a 3.0c is 3.5-3.6Ghz.

And for some people stability and compatibility is everything. I doubt i would even consider building an A64 setup for someone that truly cared about ither of those aspects.

 

[Koz]

You obviously didnt even read what i wrote. The FPS difference was 60FPS vs. 80FPS a VERY noticeable difference.

I have more faith in Hard than TechReport, as it doesnt seem to match the typical patterns in even standard benches...

http://www.aceshardware.com/read.jsp?id=60000315
Theres a good review for you to look at.
http://anandtech.com/cpu/showdoc.html?i=1941
Another informative review.

The A64 platform is totally stable and 95% of stability problems are user error, NOT hardware. Thats mostly just fanboyish accusations. Some people have stability problems during overclocking, but that is completely a user problem.
As for "average" overclocking:
http://www.cpudatabase.com/CPUdb/

Look there. Its a database of overclocks. If you look at JUST air overclocks (ignoring the extreme vapochill clocks), the 3.0C is averaging more like 3.3-3.4 Overclocked. The AMD 3000+/3200+ however is averaging 2.4GHz. Both reach aproximately a 400MHz overclock, but on an AMD processor, this means a far greater performance jump.

 

[CIWS]

Originally posted by Josh_B
Intel and AMD cross license each other... so who cares? Are you going to criticize Transmeta and VIA for using IA32 as well?

Here's the man that made the main point.

BOTH companies made an agreement to SHARE each others tech. That doesn't mean one is stealing from the other. It means both parties agreed that it was in their best interests for survival to share. Now I don't know about you, but if a company like Intel with all of it's cash and resources is willing to make such an agreement then they must feel that at least some of AMD's patents are worth having in their products. Enough that they are willing to allow them to use the tech / patents that they invent as well. If you think that Intel has not used anything AMD has invented (or at least patented first) then your fooling yourself. They BOTH do it.