AMD and Intel cpus

[[21CW]HIVictim]

Re: AMD and Intel cpus

How isn't this true? Intel developed the entire IA-32 instruction set architecture, and AMD came along and licensed it from Intel. It's like generic medications. One laboratory does all the research and all the formulating, and a smaller company comes along and makes their own cheaper version of the same medication and sells it at a reduced price.

In fact, the only real contribution AMD has made to the current architecture is towards the x86-64 portion of it. AMD made contributions to 64-bit computing in the home x86 environment, but without Intel's IA-32 architecture and instruction sets, AMD would not be in the position it currently is in.

You are comparing an architecture designed over 10 years ago. Back in those days, yeah there wasn't much difference between an AMD 700mhz and Intel 700mhz chip. But the architectures are drastically different now and there is very very little similarities between them. You cannot compare the chips 10 years ago to the chips today, it is simply ridiculous to even make a comparison of that nature.

[Busby, Paul (Bicci)]

Re: AMD and Intel cpus

How isn't this true? Intel developed the entire IA-32 instruction set architecture, and AMD came along and licensed it from Intel. It's like generic medications. One laboratory does all the research and all the formulating, and a smaller company comes along and makes their own cheaper version of the same medication and sells it at a reduced price.

In fact, the only real contribution AMD has made to the current architecture is towards the x86-64 portion of it. AMD made contributions to 64-bit computing in the home x86 environment, but without Intel's IA-32 architecture and instruction sets, AMD would not be in the position it currently is in.

64bit is a pretty major contribution though...

There is a lot more to it all than just intruction sets otherwise all CPUs would perform in a near enough identical fashion. AMD where also the first consumer desktop CPU manufacturer to create a native dual core CPU and the first to integrate a memory controller.

I know what your saying but fact of the matter is there is a lot more to an entire CPU than just the initial x86 intruction sets

[AlecJ32]

Re: AMD and Intel cpus

You are comparing an architecture designed over 10 years ago. Back in those days, yeah there wasn't much difference between an AMD 700mhz and Intel 700mhz chip. But the architectures are drastically different now and there is very very little similarities between them. You cannot compare the chips 10 years ago to the chips today, it is simply ridiculous to even make a comparison of that nature.

The architecture was designed over 10 years ago, but that doesn't change the fact that it's still in use, and not only that, but by far the most widely used architecture in desktop PCs.

Small changes in pipelines or die layouts do not reflect any change in the instruction set architecture, and any software that works on an i386 will work on a Core 2 Duo as well. Why is that? Because the instructions the CPUs handle have not changed; They're still the same ones developed by Intel which AMD is licensing rights to use.

Compare an IA-32 CPU to an Itanium CPU, or an UltraSPARC CPU, and the differences between AMDs CPUs and Intel's x86 CPUs are minimal and insignificant. It doesn't take an assembly programmer to tell you that.

There is a lot more to it all than just intruction sets otherwise all CPUs would perform in a near enough identical fashion. AMD where also the first desktop CPU manufacturer to create a native dual core CPU and the first to integrate a memory controller.

....Nope, because the architectural differences only change how quickly instructions may be executed. The hardware still has to be designed around and conform to the instructions it will be executing.

Again, back to the medication analogy. A generic version of an anti-histamine medication may contain different ingredients, but the active ingredient is still the same as that of the name brand medication, and the desired result is still the same. It's going to treat allergy symptoms, because that's what the medication's active ingredient does. The medication might not be quite as effective, but it still accomplishes the exact same thing as its name brand counter part. Even though the medication may be formulated slightly different, it's still basically the same medication, because it still based around the same active ingredient and has the same effect. Changing some of the inactive ingredients is not going to suddenly cause the medication to relieve constipation instead of allergy symptoms. The most it will do is change how effectively the medication treats allergy symptoms.

In the CPU architecture world, it's not any different. There may be small differences between the two CPUs, but when it comes down to it, the instruction set is still the same, and the code used is still the same. The CPUs accomplish the same end results.

[[Expletive Deleted]]

Re: AMD and Intel cpus

Now someone question the other man's parental lineage! Question the legality of his mother's profession!

(It worked last time...)

[[21CW]HIVictim]

Re: AMD and Intel cpus

The architecture was designed over 10 years ago, but that doesn't change the fact that it's still in use, and not only that, but by far the most widely used architecture in desktop PCs.

Small changes in pipelines or die layouts do not reflect any change in the instruction set architecture, and any software that works on an i386 will work on a Core 2 Duo as well. Why is that? Because the instructions the CPUs handle have not changed; They're still the same ones developed by Intel which AMD is licensing rights to use.

Compare an IA-32 CPU to an Itanium CPU, or an UltraSPARC CPU, and the differences between AMDs CPUs and Intel's x86 CPUs are minimal and insignificant. It doesn't take an assembly programmer to tell you that.



....Nope, because the architectural differences only change how quickly instructions may be executed. The hardware still has to be designed around and conform to the instructions it will be executing.

Again, back to the medication analogy. A generic version of an anti-histamine medication may contain different ingredients, but the active ingredient is still the same as that of the name brand medication, and the desired result is still the same. It's going to treat allergy symptoms, because that's what the medication's active ingredient does. The medication might not be quite as effective, but it still accomplishes the exact same thing as its name brand counter part. Even though the medication may be formulated slightly different, it's still basically the same medication, because it still based around the same active ingredient and has the same effect. Changing some of the inactive ingredients is not going to suddenly cause the medication to relieve constipation instead of allergy symptoms. The most it will do is change how effectively the medication treats allergy symptoms.

In the CPU architecture world, it's not any different. There may be small differences between the two CPUs, but when it comes down to it, the instruction set is still the same, and the code used is still the same. The CPUs accomplish the same end results.

I can understand that, but comparing CPU's based on the instruction set is just stupid. All CPU's do the same thing, but they are also very different.

[AlecJ32]

Re: AMD and Intel cpus

I can understand that, but comparing CPU's based on the instruction set is just stupid. All CPU's do the same thing, but they are also very different.

Instructions make more difference than anything else. When it comes down to actually working with the CPU, such as through assembly language, you won't see much difference between an Intel x86 CPU and an AMD CPU. If you're working with a PowerPC CPU, or an Intel Itanium, or a Sun UltraSPARC, you'll notice they're vastly different, whereas software compiled for an i386 CPU will generally work correctly on any AMD CPU and any Intel x86 CPU.

So unless you're sitting there putting together waffers and ICs down in Silicon Valley, those small design differences aren't going to make any real observable difference to you, aside from in the performance of the chip.

[Busby, Paul (Bicci)]

Re: AMD and Intel cpus

Now someone question the other man's parental lineage! Question the legality of his mother's profession!

(It worked last time...)

Few more posts then think of something else witty to shut us up

....Nope, because the architectural differences only change how quickly instructions may be executed. The hardware still has to be designed around and conform to the instructions it will be executing.

In the CPU architecture world, it's not any different. There may be small differences between the two CPUs, but when it comes down to it, the instruction set is still the same, and the code used is still the same. The CPUs accomplish the same end results.

But surely its these differences in architecture which create the differences in speed which make Intel and AMD so different.... in the very very most simple way of looking at it the two things do solve the same problem/accomplish the same result but saying one is a an offspin of the other is just pure crap after 10+ years of development from those initial license rights.

This isnt a meant to be a discussion about the very basics how how CPUs work and where their developements came from, its how things are now and what the differences are now. Minor may they be in the larger scale of things, confined to desktop CPU speak the differences are quite major.

[Flt. Sgt. Powderhound]

Re: AMD and Intel cpus

Instructions make more difference than anything else. When it comes down to actually working with the CPU, such as through assembly language, you won't see much difference between an Intel x86 CPU and an AMD CPU. If you're working with a PowerPC CPU, or an Intel Itanium, or a Sun UltraSPARC, you'll notice they're vastly different, whereas software compiled for an i386 CPU will generally work correctly on any AMD CPU and any Intel x86 CPU.

So unless you're sitting there putting together waffers and ICs down in Silicon Valley, those small design differences aren't going to make any real observable difference to you, aside from in the performance of the chip.

Wasn't that the point to start with?

[Busby, Paul (Bicci)]

Re: AMD and Intel cpus

Wasn't that the point to start with?

Exactly

[AlecJ32]

Re: AMD and Intel cpus

But surely its these differences in architecture which create the differences in speed which make Intel and AMD so different....

That's the thing. They're not so different. The differences in speed are not large, and are not consistent. They depend more on the tier at which the CPU is aimed than on which brand. Naturally, AMD's current best CPU is going to outperform Intel's current low end CPUs, and Intel's current best CPU is going to outperform AMDs current low end CPUs no matter which generation of CPUs you look at. There just isn't any clear difference between the two.

in the very very most simple way of looking at it the two things do solve the same problem/accomplish the same result but saying one is a an offspin of the other is just pure crap after 10+ years of development from those initial license rights.

How so? I could say cows are purple, but that doesn't mean anything unless I can back it up. So if what I said is pure crap, why don't you back that up?

This isnt a meant to be a discussion about the very basics how how CPUs work and where their developements came from, its how things are now and what the differences are now. Minor may they be in the larger scale of things, confined to desktop CPU speak the differences are quite major.

Nope. Were this true, then consistently, from generation to generation, one brand would always have the better product. Yet this isn't the case. There is a large amount of over lap between the two brands, and if one has the better product through one generation, the next generation will most likely be better for the other.

There's no difference. One brand isn't better at certain tasks than the other, and there's no consistent correlations of any kind to show any differences between the two.

When you look at CPU designs that are considerably different, I.E. an RISC CPU and a CISC CPU, there is a clear and easy to distinguish line between the two and what they excel at. For example, the RISC CPU is going to excel at tasks that can be done with simple instructions, while in tasks that require very complex instructions which need to be broken down, the RISC CPU may not perform as well as the CISC CPU.

Wasn't that the point to start with?

No, because that is not consistent enough from generation to generation. There is no strongly correlated difference between the two when it comes to performance.

[Busby, Paul (Bicci)]

Re: AMD and Intel cpus

But you have backed your point up by comparing them to non 86 CPUs and medicine... whats the point in that? You wanna see proof then go look at benchmarks and read up on the differences in the architecture, its not hard to find this info and you seem to love knowing about it all... plus yourself and i have both proved they are not just a spin off (ie dualcore and 64bit) but also remember AMD where the first cosumer desktop CPU manufacturer to release native quadcore too - is that dependand ont he original x86 instructions? - no its not

Nope. Were this true, then consistently, from generation to generation, one brand would always have the better product. Yet this isn't the case. There is a large amount of over lap between the two brands, and if one has the better product through one generation, the next generation will most likely be better for the other.

How does what i said mean that... seriously explain to me how you got from what i said to this statement because im at a loss here.

There's no difference. One brand isn't better at certain tasks than the other, and there's no consistent correlations of any kind to show any differences between the two.

Soooo i7 is no better than phemon II? and Core2 is no better than AM2? 939 was the same as pentium 4?

Some processors ARE better are certain tasks than other (again search the benchmarks)

When you look at CPU designs that are considerably different, I.E. an RISC CPU and a CISC CPU, there is a clear and easy to distinguish line between the two and what they excel at. For example, the RISC CPU is going to excel at tasks that can be done with simple instructions, while in tasks that require very complex instructions which need to be broken down, the RISC CPU may not perform as well as the CISC CPU.

Who the hell cares about all the other purpose built CPUs please will u shut up listing all these thing which are totally irrelevant to the topic... noone is impressed by your knowledge of electronics, especially me.

Basically if you think that there is no difference between AMD and Intel then stop posting in this thread and leave it to the people who care about the subtle differences which make the difference between buying one or the other.

[=98=]

Re: AMD and Intel cpus

Someone deleted my post the silly admin :p

Abj, you're so wrong it's embarrassing. Sure, the X86 instruction set is the same, but you're implying that there is no performance difference between either AMD or Intel processors?

Is that correct? You do seem good at starting these arguments :p

[Busby, Paul (Bicci)]

Re: AMD and Intel cpus

and continuing them after being proved wrong :p

[Wizrdwarts]

Re: AMD and Intel cpus

Brand doesn't matter in performance. The chip itself does. I don't care which company my CPU comes from as long as it performs well for a decent price.

[Shawn Zipay]

Re: AMD and Intel cpus

Hasn't this been what was being said in this thread the entire time? This one post is just a nice "tl;dr" for everyone it seems like.