I see absolutely no reason to go Intel now

[Xeon]

Its not about "helping intel develop" anything, that doesnt give you any rights, its about the cross licencing S3 and intel had. VIA claimed this cross licence they aqcuired through S3 allowed them to produce pin compatible cpu's; and without the original contracts and a degree in law, I don't think anyone of us can make a usefull comment on it.

No VIA may have had legal rights to the P3 bus due to the acquisition of S3 but VIA had no legal right to the P4 bus that they tried and failed to use without correct licensing. You can say we don’t have legal degrees to make a point, it won’t change the fact VIA tried to steal legitimate technologies from Intel and claim them as their own.

<A HREF="http://news.com.com/2100-1001_3-242800.html?tag=st_rn" target="_new">They loose</A> like the theives they are.

You mean like when intel didnt worry when VIA took 50% of the chipset business with its crappy Apollo Pro 133 ?

Oh great one a link please.

I guess you werent paying attention then.

Nope guess not better provide a link to these fantastic statistics you pulled.

Thats how it ended; it started with DEC suing intel for patent infringement, and they proved their case in court. It was settled, sure, doesnt make intel any more innocent than whatever Crash claims VIA is guilty off. Intel just bought its way out.

No actually DEC was being retarded since they had a cross licensing agreement with Intel as well. They wanted money plain and simple, and they got it when Intel gutted their company and acquired all their patents.

Oh be sure to <A HREF="http://news.com.com/2100-1023-279719.html" target="_new">this</A> and <A HREF="http://www.intel.com/pressroom/archive/releases/CN102797.HTM" target="_new">this</A> out seems only a few months later Digital makes the call and caves, legal dispute over, and when it comes down to it who cares what Cyrix wanted VIA absorbed them shortly afterwards too.

Regardless, once more, intel was sued for monopoly abuse, and illegally making it impossible for Intergraph to compete. though the case isnt over yet, they where already forced to pay $300M with possibly much more to come.

Ya I suppose and <A HREF="http://news.com.com/2100-7341-5157749.html" target="_new">look</A> Intel wins again, shucks eh how their claims are tossed out. Actually come to think of it Intergraph did <A HREF="http://news.com.com/2100-1001_3-978245.html?tag=st_rn" target="_new">this</A> just to add to this shady company’s history.

Tot he best of my knowledge, it does cover extentions to the x86 ISA such as SSE1/2/3, and AMD64. The two year "cooling down" period may or may not be true (I doubt it, considering SSE3 is expected this year from AMD, maybe even this quarter), so If you have any proof otherwise, I'll be glad to read it.

It is unknown if the cross licensing agreement covered SSE3 or not. But if the case is they are all covered with the agreement Ill go with the AMD can't deliver anything on time prognosis and leave it at that.

What a load of crap.

Oh oh you going to tell me that AMD64 is a pure 64bit cpu that the whole insides have been redone and redone to 64bit, remade pointers, registers(all of them not the 16 they added), 64bit AGU's and FPU's?

The core is a 32bit core with 64bit extensions added to it. The core is not 64bit the core has 64bit extensions that allow 64bit code to be run. The core is a 32bit core with 64bit extensions. Really how many times do I need to say that before it get threw your head? It’s a32bit chip pretending it’s a 64bit chip.

From reading what you have said about VIA and the lot I have to say I have a very similar opinion of you as Crashman does.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[trooper11]

well ok i see you mean the increase in fsb made the differnce to customers, I agree, ppl are mostly that blind to just follow numbers wether it is right or not. But today, taht isnt the case anymore so that argument cant be used on totday's prcessors, evne itnel seems to think a numbering scheme will help them out. if amd can get a bit more mainstream use, they might be able ot compete in that all important 'dell' marketplace. that is a major reason intel has a hold on the market.

 

[P4Man]

>Oh great one a link please.

<A HREF="http://www.realworldtech.com/page.cfm?ArticleID=RWT010500000000" target="_new">http://www.realworldtech.com/page.cfm?ArticleID=RWT010500000000</A>

During the last half of 1999, VIA enjoyed quite a bit of success with their Apollo Pro133 chipsets (Pro133 and Pro133A). Recent conversations with 5 of the top 10 motherboard manufacturers has revealed that shipments to OEMs are predominantly VIA based boards - between 60% and 70% of the total! I was actually taken somewhat aback upon first hearing this number, but subsequent discussions with others has confirmed this fact.

Regardless of precise numbers, the point stands that intel was losing substantial marketshare to VIA. Charging fees per chipset is an obvious way to counter this loss, but its really a monopoly abuse IMHO. Intel is lucky AMD is still around, so it isnt a legal monopoly.

>Nope guess not better provide a link to these fantastic
>statistics you pulled.

This wasnt about statitistic, but the fact intel used strongarm tactics to delay K7 boards.

http://www6.tomshardware.com/motherboard/19990818/athlon-01.html
http://www.theinquirer.net/?article=9139
http://www.theinquirer.net/?article=9366
http://www.anandtech.com/showdoc.html?i=1210
http://www.anandtech.com/showdoc.html?i=1392&p=2

>No actually DEC was being retarded since they had a cross
>licensing agreement with Intel as well

They got one as part of the settlement, AFAIK, they didnt have one before

>They wanted money plain and simple, and they got it when
>Intel gutted their company and acquired all their patents.

Sure, the settlement served both companies, but I don't see any evidence Digital's claims where unfounded. I can't access the intel page you linked right now however, I'll try again later.

>Ya I suppose and look Intel wins again, shucks eh how their
>claims are tossed out.

I'm not following this case very closely, but it seems like the court sided with intel on the Itanium patent infringement claim, but not on the Pentium patent infringement for which intel has already paid $300M as settlement <A HREF="http://news.com.com/2100-1001_3-882720.html?tag=mainstry" target="_new"> link </A>.

>It is unknown if the cross licensing agreement covered SSE3
>or not.

AFAIK, the deal covers x86 and all its extentions, like SSE1/2/3 but also 3Dnow and AMD64. Either way, its a fact AMD will incorporate SSE3 into their next core revision.

> But if the case is they are all covered with the agreement
>Ill go with the AMD can't deliver anything on time
>prognosis and leave it at that.

LOL.. well I guess SSE3 really is a hot issue, and AMD is scrambling to incorporate it. BTW, where are those x86-64/CT/IA32e/EM64T/ chips from intel ?

>Oh oh you going to tell me that AMD64 is a pure 64bit cpu

Yes, its a "pure" 64 bit cpu, not any less 64 bit than Power, Sparc, Mips or PA-Risc all of which have also been extended from 32 architectures and are backwards compatible. Unlike Alpha it does however also support 32 bit compatibility modes, doesnt make it any less 64 bit though.

>that the whole insides have been redone and redone to >64bit,

No.

>remade pointers,

Hu?

>registers(all of them not the 16 they added),
>64bit AGU's and FPU's?

Yes the ALU's and AGU's have been extended to 64 bits. FPU's don't have anything to do with it, they have been 80 bit ever since the 387.

>The core is a 32bit core with 64bit extensions added to it.
>The core is not 64bit the core has 64bit extensions that
>allow 64bit code to be run.

You read too much of intel's marketing crap. K8 is a 64 bit cpu, period. Its obviously based on a 32 bit core, but that doesnt change the fact it IS a 64 bit cpu, as much as any other 64 bit cpu. It has 64 bit wide registers, ALU's and AGU's. The fact it also supports 32 bit operation doesnt change that. You could theoretically disable 32 bit compatibility mode on a K8, would it then be a real 64 bit cpu ? Maybe you should not call Itanium a 64 bit cpu because it also supports 32 bit real mode x86 code ? If you don't like calling it a 64 bit cpu, call it a 96 bit cpu if you like, but try to fool someone else with your marketing BS that K8 is not a "real" 64 bit cpu and Itanium would be. The only thing that is true, is that K8 is mostly used as 32 bit chips for now, because there is not much software yet, but that doesnt change a thing about the architecture.

By your reasoning the 386 or even the Pentium4 would not even be a 32 bit cpu, since its based on the 16 bit 8086, which was based on the 8 bit 8088, and the P4 still supports those modes (at least the 16 bit one, not sure about 8bit). It even boots into 16 bit mode, should we call the Pentium 4 a 16 bit chip now ?

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Xeon]

Regardless of precise numbers, the point stands that intel was losing substantial marketshare to VIA. Charging fees per chipset is an obvious way to counter this loss, but its really a monopoly abuse IMHO. Intel is lucky AMD is still around, so it isnt a legal monopoly.

No actually the precise numbers are very important. It’s far too vague to even speculate the exact numbers. It also doesn’t change that VIA screwed with them and then go smacked off their high horse.

http://www6.tomshardware.com/motherboard/19990818/athlon-01.html
http://www.theinquirer.net/?article=9139
http://www.theinquirer.net/?article=9366
http://www.anandtech.com/showdoc.html?i=1210
http://www.anandtech.com/showdoc.html?i=1392&p=2

Oh god I knew you would pull this [-peep-] out. It's not my problem AMD couldn’t get their ducks in a row, its not my problem they didn’t have their own viable chipsets to start with, and its not my problem Intel scared the OEM's, the Vendor, and retail shops.

This is the way it is business is like war if AMD cant fight it like a real cut throat company that they pretend to be that’s not my frikken problem either.

AMD hurts AMD, with their poor investor relations, bad or dare I say horrible management, piss poor marketing, inability to supply their partners, lack of fabs, product delays, debt and whatever else this company royally sucks at.

The US government has legislation that monitors and manages situations like this. They have smacked Intel with Anti Trust suits before and will do it again to be sure. But the fact they saw nothing Anti Trust at all in what Intel did from 1998 to 2002 speaks for itself, you said it yourself actually we don’t have law degrees to say otherwise.

Sure, the settlement served both companies, but I don't see any evidence Digital's claims where unfounded. I can't access the intel page you linked right now however, I'll try again later.

It don’t frikken matter what the situation was originally Digital's claims were dismissed in court and then they caved and allowed themselves to be bought out. I don’t see the US watch dogs jumping on that deal, why because it’s all legal over here.

I'm not following this case very closely, but it seems like the court sided with intel on the Itanium patent infringement claim, but not on the Pentium patent infringement for which intel has already paid $300M as settlement link .

You wasting my god damned time look at the bloody dates the 300 million deals was in May 13, 1997, my link is for February 11, 2004 HUGE bloody difference.

AFAIK, the deal covers x86 and all its extentions, like SSE1/2/3 but also 3Dnow and AMD64. Either way, its a fact AMD will incorporate SSE3 into their next core revision.

I haven’t been able to find any supporting documents on the licensing between the two companies, so this is really all speculation at best; I would assume though that it was covered. Still doesn’t change the fact its AMD it’ll be late like usual.

LOL.. well I guess SSE3 really is a hot issue, and AMD is scrambling to incorporate it. BTW, where are those x86-64/CT/IA32e/EM64T/ chips from intel ?

How the hell am I supposed to know I don’t work there, they said next quarter then it will be next quarter.

Yes, its a "pure" 64 bit cpu, not any less 64 bit than Power, Sparc, Mips or PA-Risc all of which have also been extended from 32 architectures and are backwards compatible. Unlike Alpha it does however also support 32 bit compatibility modes, doesnt make it any less 64 bit though.

Oh Oh I knew you would say that too, Jesus what so hard to understand with it. Intel says it Microsoft says it "64-bit extensions" look at their sites look at the product briefs.

Christ with your flawed logic I can say the P3 and P4's was 128bit because of the SSE engine even though they are extensions. Why because the cores are 32bit calculating machines, they aren’t 128 bit because of the extensions.

With this in mind it's the apparent reason why the Athlon 64 and games in native 64bit mode both suck.

Also no the Athlon 64 is nothing like the other 64bit chips, Itanium, SPARC, MIPS... They are native 64bit code 100% of the time. They take 64bit instructions/data and chew at it in 64bit mode then spit it out 64bit. It’s like forcing 32bit code down the 286's throat it won’t do it. It can’t break the instructions/data down to something it can deal with. This process has been addressed with x86-64 AMD64 or IA-32e which ever you want to call it.

The K-8 and the Prescott are both natively 32bit machines, that can utilize 64bit addressing and execute specialized x86-64 instructions through 64bit extensions. There isn’t some magic way to look at this; it’s exactly the way you read it. It extends x86 to 64bit plain and simple.

Just remember real 64bit computers can process numbers that are 4.3 billion times as large as those processed by their 32-bit counterparts.

Yes the ALU's and AGU's have been extended to 64 bits. FPU's don't have anything to do with it, they have been 80 bit ever since the 387.

X87 uses 80-bit registers to do double-precision floating-point. The floats themselves are 64-bit, but the processor converts them to an internal, 80-bit format for increased precision when doing computations. With this in mind it can be said that the FPU of any 32bit machine can handle 64bit code theoretically. This isn’t the case unfortunately since the extra precision is lost.

By your reasoning the 386 or even the Pentium4 would not even be a 32 bit cpu, since its based on the 16 bit 8086, which was based on the 8 bit 8088, and the P4 still supports those modes (at least the 16 bit one, not sure about 8bit). It even boots into 16 bit mode, should we call the Pentium 4 a 16 bit chip now ?

Backwards compatibility has to be built into the core it has to remain binary compatible.

Remember the Athlon 64 and the Prescott are 32bit machines that have 64bit extensions. The Itanium, is a 64bit machine that emulates 32bit machines.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[P4Man]

>No actually the precise numbers are very important. It’s
>far too vague to even speculate the exact numbers

Okay than, 60% is your exact number. Try and see this in the context of the original discussion where you claimed this:

Also it's VIA, Intel didn't need to worry about their [-peep-] chipsets lowering their sales on chipsets

I think its obvious intel did worry and for good reason.

> and its not my problem Intel scared the OEM's, the Vendor,
>and retail shops.

Again, look at the perspective please. Crash stated this:

VIA used extortion to prevent Abit, Asus, and MSI from releasing their 735 chipset boards

I merely showed this is no different as what Intel did in 1999.

>The US government has legislation that monitors and manages
>situations like this.

Exactly, that is in fact what I am claiming; leave it to the courts to rule on things like patent infringements and or monopoly abuse. Once more, I do not try to prove how evil intel is, just how ridiculous crashmans statements on VIA where. To refresh your memory:

VIA breaks laws that we uphold with great esteem here, laws which are part of our culture, laws that make us "american". They steal ALL there new technology (Americans think of themselves as inovators) and use strong arm tactics to block other companies from doing business (WE'VE had anti-trust laws on the books for over 100 years to prevent that).

This type of crime has been around forever, it's definately not American. It might be human, but only in the basist of human nature, greed, kill them before they kill me, etc. That might go over well in certain undeveloped contries, or even Tiawan, but Americans are above that sort of thing.

Using that as an argument not to buy a VIA based motherboard is rather insane IMHO. If that is a valid argument, buying intel would be just as "unethical", and lets not even mention Microsoft. In fact, I think you'd have a hard time buying ANY product from any company.

>You wasting my god damned time look at the bloody dates the
>300 million deals was in May 13, 1997, my link is for
>February 11, 2004 HUGE bloody difference.

What does the date have to do with it ? Intel settled the case on the pentium, most likely because they would lose, and therefore most likely infringed those patents. Not a big deal to me, but not different from what crashman claims "evil via" does either.

>Also no the Athlon 64 is nothing like the other 64bit
>chips, Itanium, SPARC, MIPS... They are native 64bit code
>100% of the time. They take 64bit instructions/data and
>chew at it in 64bit mode then spit it out 64bit.

Bollocks. Sparc, MIPS, PA Risc and power all started life as 32 bit ISA's, and where extended to 64, just like AMD64. The only difference is those architectures have migrated years ago, decades in some cases, so software and OS support has followed. AMD64 is not even a year old, so most software isnt available yet. Big deal.

Initial Ultra Sparc servers also ran 32 bit OS's and software, and many still do today, as did Power and PA Risc machines, its the exact same thing. In fact, 32 bit apps are still very common on both PA Risc and SPARC, I don't know about Power, but I would be surprised if it where any different. To give you an example, where I work we do a lot of development using Speedware on HP-UX on PA Risc hardware. Speedware is not even available as a 64 bit port, and there are no plans, because there is no use for it. The 32 bit version works just fine. Does that mean PA Risc is a 32 bit ISA now ? You do know HP-UX only became available here as a 'mostly' 64 bit OS in 1998 (v11) ? Most companies didnt even touch it until 2000, many of them still run 32 bit HP UX. most software today is STILL 32 bit, either under 32 bit or 64 bit HP-UX.

>The K-8 and the Prescott are both natively 32bit machines,
>that can utilize 64bit addressing and execute specialized
>x86-64 instructions through 64bit extensions. There isn’t
>some magic way to look at this; it’s exactly the way you
>read it. It extends x86 to 64bit plain and simple.

Like I said, you read too much intel PR. Tell me this, what is your definition of a 64 bit cpu ?

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Xeon]

Okay than, 60% is your exact number. Try and see this in the context of the original discussion where you claimed this:

Hmm 50% of the top 10 motherboard manufacturers sold 60%-70%. Its unknown what the other players sold do your numbers are invalid. You can't get any even vague % because we don't know the sales of the other company’s hell they could have been 1% of their sales and that would totally skew the results.

I merely showed this is no different as what Intel did in 1999.

Oh so Intel strong armed them and guaranteed sales for the manufactures, something AMD could not. While VIA just tried to kill any other 735 chipsets till theirs was ready and had higher performance.

Using that as an argument not to buy a VIA based motherboard is rather insane IMHO. If that is a valid argument, buying intel would be just as "unethical", and lets not even mention Microsoft. In fact, I think you'd have a hard time buying ANY product from any company.

Don't buy them your European I am sure they have good stuff being made over there *giggles*.

Also I will rehash what I said before its business and the US government deals with it in accordance. Intel is big and mean I never denied that. But VIA has danced with Intel for 15 years at least, and every time Intel has slammed them back into their jungle hole.

Well to be more specific VIA's CEO has been pestering Intel for a long time.

What does the date have to do with it ? Intel settled the case on the pentium, most likely because they would lose, and therefore most likely infringed those patents. Not a big deal to me, but not different from what crashman claims "evil via" does either.

Those were the initial Intergraph suit for Itanium patent infringements. The second was a repeal of that suit in Intel’s favor. I then showed another link where Intergraph is a second rate chump company and was trying to sue for everything under the sun because they are a failing company.

The only difference is those architectures have migrated years ago, decades in some cases,

You said it yourself there was a complete machine move over form 32bit to 64bit. This case actually is similar to AMD's since I never noticed the AGU ports had been increased to 8byte widths. But it still remains that the core is a native 32bit core.

Reason being is when your running 32bit mode you can add 64bit mode in as well, you can access those extra registers and gain the extra performance not inherit to 64bit but to the additional code being decoded/executed/and retired.

This is not the case in 64bit mode the CPU is unable to access the 32bit registers only the SSE/SSE2 registers in fact since they deal with native 64bit code abet specialized in nature. From here we see the weakness of the design's Prescott included.

While the Itanium is always called into play when it comes down to this, it doesn’t have any real x86 execution engines in there all it has is a emulator that converts x86 code on the fly to IA-64 code.

So I will restate it again x86-64 is still limited to x86 instruction messiness in the 64bit mode. Why because all the 64bit is, is the 32bit commands/instructions what ever you want to call them redone to support 64bit. There is no new ISA there is plain Jane x86 with 64bit capabilities added.

Like I said, you read too much intel PR. Tell me this, what is your definition of a 64 bit cpu ?

Something that has to emulate x86, something that natively runs in 64bit mode, something has been 100% designed around 64bit specifications not some hack job where they widened pathways.

Also regardless of how the other 64bit competitors started out they all moved to a new ISA (some did leave 32bit backwards capability in there such as Alpha), but in the end they were rebuilt from the ground up to be 100% compliant to 64bit.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[darko21]

Re: Something that has to emulate x86, something that natively runs in 64bit mode, something has been 100% designed around 64bit specifications not some hack job where they widened pathways. Also regardless of how the other 64bit competitors started out they all moved to a new ISA (some did leave 32bit backwards capability in there such as Alpha), but in the end they were rebuilt from the ground up to be 100% compliant to 64bit.

Spud let it go you are just grasping at straws and digging a deeper hole everytime you reply.

Re: Something that has to emulate x86

By this definition no x86 cpu could ever be 64bit if you feel that way how could a 16bit cpu become a 32bit cpu?

Re: something that natively runs in 64bit mode, something has been 100% designed around 64bit specifications not some hack job where they widened pathways.

You mean like itanium? have you ever given any real thought to why itanium does not run 32bit code. or how about why itainium don't run 16bit code.

Re: Also regardless of how the other 64bit competitors started out they all moved to a new ISA (some did leave 32bit backwards capability in there such as Alpha), but in the end they were rebuilt from the ground up to be 100% compliant to 64bit.

Look at it this way take the k7 or p3 they don't run x86 in emulation now do they. You call them 32bit cpu's yet they run 16 bit code so why do they deserve to be called 32bit cpu's you could use you silly arguments to claim a p3 is not a 32bit cpu but only a 16bit cpu.

Emulation has zero to do with it.. emulation is for running non compatible software through emulation software. like running nintendo on x86 with a big hit in performance.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[kinney]

VIA sucks.
But whats up with the A64 bashing?
Its a great cpu and obviously their first entrance into the consumer 64bit market has made them a force to be reckoned with (if intel is adopting AMD64).
While still being able to play on the 32bit playground with Intel.

You are correct, I mean it isnt a pure 64bit redesign like Itanium but being first on market makes it more viable to become the standard bearer.
And it apparantly has, or will.
Its just a smarter way of doing it, rather than building a 64bit processor and emulating 32bit like Intel, AMD build UP from 32bit to 64...
the only one in this equation that ISNT a true 32bit is the Intel method.
As intel emulates 32bit.
AMD does not "emulate" 64bit processing. Just happens to be built on a great 32bit core.

While both are "true" 64bit designs.
Correct me if I'm incorrect, by all means I'm here to learn.

VIA had its day, and will continue on.. the global market is big enough to allow crap to sell in moderately significant quantities.
I suppose I dont get what the core of your guys' pissing match is about.

That VIA did penetrate the market at one time? Yeah.. so? Back into their little Taiwanese corner they go.
They cant engineer on par with the Nvidia corporation.
Thats been displayed many times now.

I could pull out rather impressive numbers when NV jumped in the AMD chipset market how they absolutely TROUNCED VIA's punk arse.

Only fools buy the crap over a Nvidia, SIS or AMD chipset.

----
RESIDENT FORUM WARRIOR
Support the terrorists, vote democrat

 

[Xeon]

By this definition no x86 cpu could ever be 64bit if you feel that way how could a 16bit cpu become a 32bit cpu?

No a 32bit CPU is 32bit because it runs natively in 32bit. The only reason x86 can run 16bit and 8bit is because of binary backwards compatibility, its one of the more annoying features of the ISA.

Furthermore the entire internals of the machine regardless of its bit calculation design are built around it. A 32bit CPU will have a 32bit ALU and a 64bit FPU but in regards its 80bit since they want extra precision to be available. The fact that the Opteron has a 64bit ALU but a 80bit FPU still makes it a 32bit system if it was built around 64bit it would be 160 bit thus to allow for extra precision, this is not the case.

By this definition no x86 cpu could ever be 64bit if you feel that way how could a 16bit cpu become a 32bit cpu?

No because x86 is built around backwards compatibility. But it doesn’t matter what x86 is it could evolve to 128bit but it is still limited to register limitations and code structures, retire limits, table sizes, loop sizes, and decode steps to name a few. The fact the machine must be built around the bits makes it a 32bit machine or a 64bit machine.

You can't say otherwise since the case would hold that x86 has been 64bit since MMX was incorporated. It’s a marketing twist plain and simple.

You mean like intanium? have you ever given any real thought to why intanium does not run 32bit code. or how about why intainium don't run 16bit code.

Actually I know the exact reason why the Itanium can't decode and execute x86 instructions/code, without a specialized execution unit. ISA is radically different and compatibility was not even considered when Intel started development.

Look at it this way take the k7 or p3 they don't run x86 in emulation now do they. You call them 32bit cpu's yet they run 16 bit code so why do they deserve to be called 32bit cpu's you could use you silly arguments to claim a p3 is not a 32bit cpu but only a 16bit cpu.

Its part of the ISA it carried that luggage from its 8bit days it can’t be helped. They are also designed to natively run 32bit code without having to run it into protected mode, which is necessary to run in 16bit mode.

Emulation has zero to do with it.. emulation is for running non compatible software through emulation software. like running nintendo on x86 with a big hit in performance.

Wow your right hence they are different ISA's hence the need for emulation which always cost system ticks because the instructions/code have to be broken down into something that the machine its is running the emulation can deal with.

But whats up with the A64 bashing?

No Opteron bashing just AMD bashing, I have no qualm admitting the A64 beat Intel’s offerings like a red headed step child. I have no problem admitting AMD did a very good job with the technologies within especially with limited capital and R&D resources. I can’t praise the company enough for the technology that they have delivered to us. I do have a problem with the time, I do have a problem with the pricing, and I do have a problem with the lack of compilers (I know there are compilers out there, but this should really come down to the manufacture to ensure that a smooth transition can occur.)

You know what else I like even more though is the panic AMD has put Intel into, the drops in prices the performance barriers we are shattering now, the technological leaps we are experiencing. I have to thank AMD for that, but this doesn’t change my thoughts on AMD as a company, I am hoping you are able to distinguish the differences.

Only fools buy the crap over a Nvidia, SIS or AMD chipset.

Couldn’t have said it any better than that bud.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[jrosenst]

Well, if you’re in an environment where money is no object (military) and you just want the fastest thing, then atm it's Intel.

Also, for rendering, or complex math calculations (Like blast analysis) that might take 3-4 days to run, you don't care much about hardware cost, you just want the fastest thing out there.

If a top scientist is costing you 100 bucks and hour, the computer expenses are in the noise.

 

[darko21]

Re: Well, if you’re in an environment where money is no object (military) and you just want the fastest thing, then atm it's Intel.

Not if you want to run x86 stuff. what pecent of software out there is x86 if you removed the mac from the equation? what is mac anyway 5% and last time I checked intel cpu's don't support mac unless through emulation. I guess you must mean ia64 hmmmm.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart. <P ID="edit"><FONT SIZE=-1><EM>Edited by darko21 on 03/18/04 09:04 PM.</EM></FONT></P>

 

[darko21]

My god do I have to break that down...... give me a break. You should know better than that. Don't let cpu preferences get the better of ya.

the a64 is every bit a 64bit cpu as the pentium was a 32 bit cpu. PERIOD....

I'll hope p4man fills in the blanks a little better for ya. why because he knows more than me on this topic and I'm running out of energy plus it's getting boring.

All I can gather from your last post is you are saying no cpu that runs x86 could posibly be considerd 64bit. now do you really want to stand by that statment.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart. <P ID="edit"><FONT SIZE=-1><EM>Edited by darko21 on 03/18/04 09:19 PM.</EM></FONT></P>

 

[Xeon]

Then the Pentium 4 is 128bit plain and simple, since the SSE2 engine is 128bit. Also BB or P4Man as he is called now doesn't fill in any blanks all he does is gabber on about how x86-64 is 100% 64bit.

It's not and never will be its a x86 extension just like SSE/SSE2 it's nothing more and certainly nothing less. It's x86 trying to be 64bit with the disavantages that x86 brings to the table as well.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[darko21]

Re: Then the Pentium 4 is 128bit plain and simple.

No that's just stupid spud, how much virtual memory can the p4 with sse2 address?

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[Xeon]

Unknown limitations since SSE2 is a execution unit not a pointer.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[darko21]

so this makes it a 128bit cpu?

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[Xeon]

Does a 32bit CPU with 64bit extensions make it a 64bit CPU or a 32bit CPU with 64bit extensions?

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>

 

[endyen]

When is a backwards compatable 64 bit cpu not 64 bit? When it's not by Intel. Well Spud, IA64 is yours. Keep it, nobody else wants it.

 

[jrosenst]

I wrote: Well, if you’re in an environment where money is no object (military) and you just want the fastest thing, then atm it's Intel.

Then darko21 Wrote: Not if you want to run x86 stuff...

OK, so if I had unlimited funds to purchase a computer and only cared about speed and that it must be x86 compatible. Let’s say I am writing a windows app with visual studio.net and I want to do a lot of complex calculation that will take days to run. Would that PC have an Intel or an AMD processor in it? I say it will have an Intel. Give me a link that say's otherwise.

 

[darko21]

Re: OK, so if I had unlimited funds to purchase a computer and only cared about speed and that it must be x86 compatible. Let’s say I am writing a windows app with visual studio.net and I want to do a lot of complex calculation that will take days to run. Would that PC have an Intel or an AMD processor in it? I say it will have an Intel. Give me a link that say's otherwise.


Ok you want to do complex calculations or algorithum compression then Intel has a slight advantage at the moment. but you never said what platform, programs you wanted to concentrate on. just calling intel faster in general (which is what you made it look like) is wrong. yes intel has certain advantages in certain situations I wonder if that will be true once 64bit windows is out and about.

Re: Give me a link that say's otherwise.

You brought it up so back it up.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[endyen]

He's talking out of his hat. You cant use SSE2 for writing progs. You are stuck with pure ALU/ FPU so an Amd rig is the only way to go.

 

[darko21]

Re: Does a 32bit CPU with 64bit extensions make it a 64bit CPU or a 32bit CPU with 64bit extensions?

8 bit 16bit 32bit 64bit when do these x86 cpu's become pure in your mind? no wait you answerd that but it made no sense. you have been making all kinds of flakey excuses like I said the a64 is every bit a 64bit cpu like the pentium was a 32bit cpu.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[endyen]

If you want a really good laugh, see if you can find spuds prescott predictions. Reality has no place in some people's minds.

 

[darko21]

Actually I like spud I just think he has been consumed by the dark side. I will teach spud how to overcome this evil force.

USE THE FORCE SPUD USE THE FORCE.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[Xeon]

Unknown was the increase in pipeline length, otherwise those predictions would have been pretty close to reality at 20 stages.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>