Is AMD faster than intel

Quote:

---

Originally posted by joerg


Now to moridin what do you mean by leagcy application are we talking about unreal2k3 or what because as far as im concerned legacy means programs designed for an OS prior to 98. So i guess most modern games must be leagcy games then becasue the Athlon clock for clock kicks some serious butt.

---

The P4 optimization guide was not even available until 2000, so any application that started development prior to that isn’t likely to have much in the way of P4 optimizations. After that it still took time for developers to learn for themselves. It’s only been in the last the last year that we have even begun to see developers get anywhere near the capabilities of the P4.

Even worse for the P4 is that even 2-3 years ago many developers were using the Pentium classic (P5) as their optimization target. The P4 is VERY bad on code scheduled for the P5. Of course this will only show up in benchmarks, not real life since the fastest P5 was something like 233 MHz, so even the slowest P4 is plenty fast for this code.

This pattern isn’t new BTW, the predecessors of the P4 followed exactly the same pattern. The P5(Pentium classic) , and P6 (Ppro, PII, PIII) were slammed for the very same issues early in their life.

So basically what your sayign is that P7 or P4 is only now getting getting applications that are suing its matured code mainly sse2 instructions and that applications wriitten with such codes are the only non leagcy applications and games out there. So will you say the same thing abotu the new featyrse being added in hammer or will you proclaim them useless? Also once real world apps start to use sse2 AMD will have already added support for these instructions so in common apps and games Athlon will not fall behind at all. And again once more games and apps take advatage of HT i guess the same can be sayd. So maybe AMD is a parasite or maybe Intel has sofware compnaies on a leash and AMD just kinda supports what they do. The whole point of what i am saying is that Both companies will release code that can be optimized for but Intel is the company that has more chnce of seeing its instructions oprimized for because they have more of the market share, but as soon as the code becomes relavant AMD seems to always have adapted it in some form anyway so what we see in the most up to date bench marks can not be used to reflect the current state of software performance.

Quote:

---

Originally posted by kirkyg


Two things i would ask...WHERE THE HELL! did you get those ******** benchmarks? and then why dont you take a look at what benchmarks the athlon wins in. All the important ones.

kirkyg

---

dude, you need to research before mouthing off like that. and i doubt there are much people in here who can hang with Moridin on processor debate. he is the authority on Sharky's about this stuff.

Spec is one of the most important and well known benchmarks out there, as Moridin pointed out. not a silly benchmark like 3dmark, which, i admit, is fun to increase your score (i fall into this every once in a while) but definitely not important.

Quote:

---

Originally posted by steppy76
Man, I'm so tired of this worthless argument. You want to compare at equal price points, fine. But just simply testing them at the same clock does absolutely nothing but pit a strength of the AMD against a weakness of the P4... Let's just retire the "clock for clock" argument.

---

Worthless?

Even today we hear INTYEL officials lean on the "clock speed" factor/argument. So, let's humour them, let's compare chips speed for speed. By using INTEL's own arguments, the community knows that INTEL chips are easily beatable. What INTEL expects is simply this: maintain a MHZ (GHz) lead, that way it will stay ahead (even if by a slight margin) of AMD.

INTEL users/fans always resort to this type of reasoning, they seemingly dislike reality. I understand that, but face it, if people compare the 3GHz INTEL with the 3000+ XP, what lead does INTEL have here? Not much. Now, imagine if INTEL stood still and waited for AMD to catch up in terms of raw clockspeed: does anyone doubt that even a "thin" 2.7 GHz AMD would beat a "robust" 3GHz INTEL?

Chronus

Quote:

---

Originally posted by Chronus


Worthless?

Even today we hear INTYEL officials lean on the "clock speed" factor/argument. So, let's humour them, let's compare chips speed for speed. By using INTEL's own arguments, the community knows that INTEL chips are easily beatable. What INTEL expects is simply this: maintain a MHZ (GHz) lead, that way it will stay ahead (even if by a slight margin) of AMD.

INTEL users/fans always resort to this type of reasoning, they seemingly dislike reality. I understand that, but face it, if people compare the 3GHz INTEL with the 3000+ XP, what lead does INTEL have here? Not much. Now, imagine if INTEL stood still and waited for AMD to catch up in terms of raw clockspeed: does anyone doubt that even a "thin" 2.7 GHz AMD would beat a "robust" 3GHz INTEL?

Chronus

---

Whats does it matter if the P4 3.06 is clocked higher than the XP 3000+ the P4 still cost less and overclock better. So for price/performance Intel wins in the show down.

Quote:

---

Originally posted by crashed13


Whats does it matter if the P4 3.06 is clocked higher than the XP 3000+ the P4 still cost less and overclock better. So for price/performance Intel wins in the show down.

---

Though I have to admit (after a new look at newegg and http://freespace.virgin.net/m.warner/CPUPricesAMD.htm)
that INTEL's top is better in terms of performance/price (what a change!), I have to point you back to the initial post of this thread and some of the comparisons that followed.


INTEL is faster, yes. Not by a great margin, but faster it is.

But if we were to compare equal speeds cpus, INTEL would/will lose.

Now, that's a fact.

Though not really important, unfortunately for AMD right now.

Chronus

Quote:

---

Originally posted by Chronus


Though I have to admit (after a new look at newegg and http://freespace.virgin.net/m.warner/CPUPricesAMD.htm)
that INTEL's top is better in terms of performance/price (what a change!), I have to point you back to the initial post of this thread and some of the comparisons that followed.


INTEL is faster, yes. Not by a great margin, but faster it is.

But if we were to compare equal speeds cpus, INTEL would/will lose.

Now, that's a fact.

Though not really important, unfortunately for AMD right now.

Chronus

---

I know you like AMD but give it a rest already. You sound like a poor sport, when you give Intel credit for being faster and then say but if we compare equal clockspeeds its different. It seems to really bother you AMD guys that Intel's chips seem to run several hundred mhz higher. AMD's chip design is not the only way and Intels is not the wrong way. It doesnt make one bit of difference how the chips are designed to run as long as they accomplish the same thing, who cares? Even if one is faster than the other.

Quote:

---

Originally posted by crashed13


Whats does it matter if the P4 3.06 is clocked higher than the XP 3000+ the P4 still cost less and overclock better. So for price/performance Intel wins in the show down.

---

But then you payed $560 bux or so for your processor and i payed $93 shipped from newegg and my processor is faster in most benchmarks than the 3.06 DOH...who wins again?

kirkyg

Quote:

---

Originally posted by deeznuts


dude, you need to research before mouthing off like that. and i doubt there are much people in here who can hang with Moridin on processor debate. he is the authority on Sharky's about this stuff.

Spec is one of the most important and well known benchmarks out there, as Moridin pointed out. not a silly benchmark like 3dmark, which, i admit, is fun to increase your score (i fall into this every once in a while) but definitely not important.

---

I dont need to do "research" i have nothing to prove in here. All im saying is in any games that you run (besides quake3) an equivalently clocked athlon system will DESTROY the p4...there isn't even an argument there. And your on here saying all these benchmarks prove that they are really close even clock for clock . But in real world application THEY ARE NOT period.

kirkyg

Colossus,

I have a question regarding your Asus P4S533 system running at 160x21. I have the same motherboard and am just curious what "160x21" means. I presume it's 160 FSB, but 21? I don't have that option in my BIOS.

The highest I can get my system to boot at is 137/34, but it's not stable. I run it at 133/33 with memory at 4:5.

Anyway if you could help me by explaining the 160x21 thing I would very much appreciate it.

Sorry this message is unrelated to the thread...

Quote:

---

Originally posted by kirkyg


I dont need to do "research" i have nothing to prove in here. All im saying is in any games that you run (besides quake3) an equivalently clocked athlon system will DESTROY the p4...there isn't even an argument there. And your on here saying all these benchmarks prove that they are really close even clock for clock . But in real world application THEY ARE NOT period.

kirkyg

---

Your still not getting it. They are saying if the code is optimized properly the P4 can and will compete with the AMD which you fail to realize.

Quote:

---

Originally posted by kirkyg


I dont need to do "research" i have nothing to prove in here. All im saying is in any games that you run (besides quake3) an equivalently clocked athlon system will DESTROY the p4...there isn't even an argument there. And your on here saying all these benchmarks prove that they are really close even clock for clock . But in real world application THEY ARE NOT period.

kirkyg

---

first you say in "any game," and then you go on to say "real world application." games and real world appz are not synonymous in my opinion. yes a significant number of us in here play games, but the real world does not. don't get it twisted.

"And your on here saying all these benchmarks prove that they are really close even clock for clock . But in real world application THEY ARE NOT period."

SPEC is made up of real world applications as Moridin pointed out. 26 i think it was. so you are discounting this benchmark, made up of 26 real world applications, and countering it with "in any games you run." not a strong argument.

cpu2000 programs
http://www.spec.org/cpu2000/CINT2000/
http://www.spec.org/cpu2000/CFP2000/
so games are more real world huh?

I guess there isnn't much thats optimized for it...again in real world application it doesn't outperform the AMD chips it underperforms them clock for clock there is absolutely NO argument suggestioning otherwise. There are a few things that p4 does exceptionally and amd the same...but excluding those the chips are not created equally clock for clock.

kirkyg

Quote:

---

Originally posted by joerg
So basically what your sayign is that P7 or P4 is only now getting getting applications that are suing its matured code mainly sse2 instructions and that applications wriitten with such codes are the only non leagcy applications and games out there.

---

Just to be perfectly clear I'll repeat this. THE MOST IMPORTANT P4 OPTOMIZATIONS HAVE NOTHING TO DO WITH SSE2! Code scheduling, L/S rules, branch rules and data structure are far more important. These are the things I have been referring to. It takes time for programmers and compiler writers to figure out how to make these things work in their programs, and this learning curve only started in late 2000 so we have yet to see the results.

Hammer will not see the same types of benefits. AMD is forced by their market position to sacrifice top end performance in favor of good performance on a wide variety of code.

This means that over time we see better performance out of Intel's processors as software catches up. We typically do not see this in AMD processors.

Quote:

---

Originally posted by joerg

So will you say the same thing abotu the new featyrse being added in hammer or will you proclaim them useless?

---

Some of the new features in Hammer like more architectural registers in 64-bit mode are potentially very useful. But, 3DNow! was potentially useful as well, in practice it didn't turn out to be particularly important because everyone chose to use SSE instead.

Quote:

---

Originally posted by Chronus


Worthless?

Even today we hear INTYEL officials lean on the "clock speed" factor/argument.

---

Do you have some links to back that up?

Quote:

---

Originally posted by kirkyg


But then you payed $560 bux or so for your processor and i payed $93 shipped from newegg and my processor is faster in most benchmarks than the 3.06 DOH...who wins again?

kirkyg

---

Sure, but then O.C. the P4 to like 3400, pair it with Rambus and you're done:p ... You have to compare apples to apples. You also have to look at the applications for which you purchase your rig. Your 2100 couldn't even dream of running within HALF the fps my wife's p4 could do encoding mpegs and avi's. With SSE2 and HT, FORGET IT... go back to games, dude...:cool:

Quote:

---

Originally posted by kirkyg


I dont need to do "research" i have nothing to prove in here. All im saying is in any games that you run (besides quake3) an equivalently clocked athlon system will DESTROY the p4...there isn't even an argument there.

---

All that proves is that those particular games perform better clock for clock on the Athlon. It says nothing about what the two processors are capable of clock for clock, so they simply do not back up you position that the Athlon is faster clock for clock. These are application benchmarks, and application benchmarks CANNOT tell you how a processor performs in general.

Most of these games were started before P4 optimization guidelines were known, and even for those started after, there is no guarantee that the developers took the P4s needs into account. It is therefor not surprising that the P4 performs poorly on them clock for clock, This tells us nothing about what we will see in the future and it doesn’t matter anyway since the P4 still offers performance as good or better then any other processor out there due to it's high clock rate.

Quote:

---

Originally posted by kirkyg

And your on here saying all these benchmarks prove that they are really close even clock for clock . But in real world application THEY ARE NOT period.

---

You have already been pointed to real world applications that show this isn't true.

Quote:

---

Originally posted by kirkyg
I guess there isnn't much thats optimized for it...again in real world application it doesn't outperform the AMD chips it underperforms them clock for clock there is absolutely NO argument suggestioning otherwise. There are a few things that p4 does exceptionally and amd the same...but excluding those

---

All the SPEC components are real applications, and some like gzip are quite common. From Tomshardware reviews mp3 Maker platinum runs faster clock for clock on the P4 (and show no FSB/memory dependence). So do the Lightwave reviews from many sites.

Benchmarks that favor the P4 on a clock for clock basis are certainly not in a majority, but neither are they hard to find.

Quote:

---

Originally posted by kirkyg

the chips are not created equally clock for clock.

kirkyg

---

No two micro-architectures are. There are simply to many issues to address, but:

The P4 offers higher throughput of integer instructions, more loads and stores, lower latency cache, better branch predictions, lower latency for simple Integer instructions, better prefetch, deeper OoOE capabilities…

The Athlon has higher X87 throughput, dual ported L1, larger L1, suffers less from a branch mispredict, suffers less from a L/S rule violation, higher sustained issue rates, lower latency for most X87 instructions, lower latency shifts…

All of these effect how much a processor can do per clock, and each processor has it's advantages. How then can you make such sweeping statements that the P4 simply isn't capable of competing clock for clock?

you waste all that time microanalyzing everything i said trying to make a point and ended up proving mine thanks bro ;)

Quote:

---

All that proves is that those particular games perform better clock for clock on the Athlon. It says nothing about what the two processors are capable of clock for clock, so they simply do not back up you position that the Athlon is faster clock for clock.

---

Indeed they do...because the rare times that your processor is maxed out 100% resources pumping full capacity is when your playing a game or doing some sort of graphics editing/viewing....thats the ONLY time your processor is maxed. Just proved my point. In real world when you have your system maxed out when you really need the power the AMD clock for clock will DESTROY p4 for the same clock. I would buy a 3.06 over a 3000+ because it has hyperthreading and is slightly faster than the 3000+ for the same price. But the 2100+ is EASILY a better buy than anything p4 has to offer on the cheaper end.

kirkyg

Quote:

---

Originally posted by Moridin

All of these effect how much a processor can do per clock, and each processor has it's advantages. How then can you make such sweeping statements that the P4 simply isn't capable of competing clock for clock?

---

Well, the P4, without SSE2, simply isn't capable of competing clock for clock with the Athlon XP. That's quite clear from the 1000s of P4 reviews. Lets be real here though, who buys a P4 at 2.2 hoping it will beat an XP 2800+?

The P4 has to have higher clock speed to be overall "faster" than an Athlon. How a P4 performs on an SSE2 written benchmark or program that is hardly commonplace, such as Lightwave or MP3 Maker, probably doesn't mean anything to 99% of us here either.

Quote:

---

Originally posted by gotmonopoly?


Well, the P4, without SSE2, simply isn't capable of competing clock for clock with the Athlon XP. That's quite clear from the 1000s of P4 reviews. Lets be real here though, who buys a P4 at 2.2 hoping it will beat an XP 2800+?

The P4 has to have higher clock speed to be overall "faster" than an Athlon. How a P4 performs on an SSE2 written benchmark or program that is hardly commonplace, such as Lightwave or MP3 Maker, probably doesn't mean anything to 99% of us here either.

---

Thats kind of a lame arugment. Saying well if the P4 didnt have sse2 it wouldnt be capable of competeting. Thats like saying well if the athlon didnt do more work per clock cylcle then it wouldnt have a chance at competeting. Thats a pointless argument since, it does have it. Btw as Moridin has pointed out the most important optimizations for the P4 has nothing to do with SSE2.

Unless I missed something, isn't it clear that the 3.06 is the fastest chip out there. :confused: As far as all this real world stuff, I think the only thing that matters is what you do. All I can say is AMD is good for what they are, a company who has the best prices and offers great performance at those prices. :) 2100+ @ 2.4GHz for $90 can't be beaten. :cool:

Yes it is as simple as I play unreal2k3 and the athlon is awsome at that game or I use lightwave, and 3dsMax and the P4 kicks but in these applications. The simple point is that from a practical standpoint for what most of us and the rest of the world do the Pr ratings of the athlon will hold up against the P4 of course some times this isnt true but for 97% of us this will remain true throughout our cpus life. Again once enough applications are written to be optomized for the p4 then AMD will have already stepped up and gained many of these same things and again someone will make the initial move and software companies will again catch up. Oh and kirky dude games are nto the only thing that will bring a cpu to its knees. Have you ever used intensive CAD programs or rendered a movie or did and an advanced filter in photoshop how about running setti or simply crunching numbers or being a server. All of those things can bring a computer to its knees. Games wont even fully bring a cpu to 100% because the graphics card is doin most of the work just because your taskmanager says cpu is at 100% doesnt mean crap it just means every clock is being used not that every clock is being used ot its max which is why when sc4 lists at 100% and sc2k lists as 100% you can clearly say sc4 is actually doing more.

Quote:

---

Originally posted by kirkyg
you waste all that time microanalyzing everything i said trying to make a point and ended up proving mine thanks bro ;)



Indeed they do...because the rare times that your processor is maxed out 100% resources pumping full capacity is when your playing a game or doing some sort of graphics editing/viewing....thats the ONLY time your processor is maxed. Just proved my point. In real world when you have your system maxed out when you really need the power the AMD clock for clock will DESTROY p4 for the same clock. I would buy a 3.06 over a 3000+ because it has hyperthreading and is slightly faster than the 3000+ for the same price. But the 2100+ is EASILY a better buy than anything p4 has to offer on the cheaper end.

kirkyg

---

I demonstrated that on properly optimized code the P4 performs similar to the Athlon clock for clock. How dose that support your point in any way? As for "microanalyzing" all I have done is take you unsubstantiated claim and refute it. You have yet to provide any evidence to counter that.

No they don't because the games you are pointing to are based on game engines who's development began before P4 optimization guidelines were available, so they couldn't possibly be taken into account. This is why application benchmarks CANNOT provide evidence for your claim. By definition an application benchmark cannot give you any information beyond that specific application using that specific data. (If you have questions reguarding what benchmarks can and cannot tell you feel free to ask)

I think the lightspeed results are clear indication that the Athlon does not necessarily "destroy" the P4 clock for clock in processor intensive graphics applications.

P4 3.06=$543 XP 3000+=$555(No HS/fan)

P4 2.8=$345 XP 2800+=$360

P4 2.66=$233 XP 2600+=$246

P4 2.4=$160 Xp2400+=$130


Price is the same and which kind overclocks better?,,do we even have to go there?