Pentium 4 or AMD 64?

superbooga said:
Sorry, I think you need a bit of training in computer architectures. If your logic was correct, then the Athlon XP would be smoking the P4 also, since the XP's pipeline is pretty much the same as the 64. The P4 beats the XP very handily in Doom 3, and Quake 3 (The AMD 64 beats the P4 in both Doom 3 and Quake 3).

The reason the 64 wins is because of uses an internal memory controller, rather than the traditional northbridge architecture. This means memory latencies are extremely low. For example, every dram access on the P4 might take 200 CPU cycles. However, it might only take 80 on an Athlon 64. That's a huge difference in efficiency.

Since the P4 has to wait longer to access dram, cache plays an important role in performance. Cache reduces the times dram has to be accessed. However, since the 64 accesses dram so much faster, it doesn't really need a huge cache.

A P4 EE beats the Northwood by about 15% in Doom 3, so yes, that L3 cache is used. Programs and apps don't decide if the L3 cache is used or not, fundamental computer architecture dictates that it will always be used! Now whether or not it's usage will make a substantial difference in performance depends on the app.
Click to expand...

First of all everybody knows that the AXP line was supposed to go up against the P4 "A" and "B" procs not the P4 "C", "E", "EE" that your talking about. The A64 goes up against the P4 "C" and "E" while the AFX goes up against the P4 "EE" (hope your still with me ;) )

Secondly before you start to throw around dumbass comments about not knowing the fundatmentals about cpu arc. know that I am A+ cert. and that I am CSC major (thats all that you really need to know...for now) and that no shit the A64 has an integrtated memory controller :eek: So please dont try to lecture me about it nOObie

And where did you get this info about the P4 EE beating the P4 C and E by more than 15%? They used a P4 EE at 3.4Ghz but they didnt use any P4 "C" or "E" at 3.4Ghz? Just wondering what website you got those numbers from.
 
I just wanted to chime in and give you guys a little heads up on this whole A64 vs P4 gaming scenario in case you didn't realize it already.

The CPU has little to do with the performance in GPU intensive games, which the majority of games are, unless you are CPU limited.

In today's games like Far Cry, UT2004, DOOM 3, and so on, your NOT going to be CPU limited at higher resolutions with AA + AF enabled. You are GPU limited. At 1600x1200 or even 1280x1024 with high AA + AF, the P4's run just as fast as the A64's. The P4 rigs have even been beating the faster A64 setups in quite of few of the recent CPU scaling reviews. This could have to do with the simple fact that Intel has better chipsets and drivers.

If you will look at Hards DOOM 3 benchmarks you will see that the Athlon FX-53 with the 6800GT is only 1 fps faster then the 3.0c and 6800GT.

Gaming is a poor reference for a CPU buying descision unless you get your jollies off by sitting around playing at 800x600.

When you look at CPU performance you need to be worried about CPU intensive tasks. Not GPU intensive tasks like Gaming. There are only a handful of truly CPU intensive games out there such as Elder Scrolls III: Morrowind.

CPU intensive tasks include: Multitasking, Rendering, Compression, Encodeing, Productivity Software, Math Analysis, and so on. The Pentium 4 architecture is faster in the majority of these areas, especially with HyperThreading in multithreaded environments.

Anandtech just put up a new Article benching one of the new Xeon EM64T chips against an A64 3500+ in Linux 64-bit and the A64 litterally got its ass handed to it in math intensive applications.

http://www.anandtech.com/linux/showdoc.aspx?i=2158

The only real advantage of an A64 is the on-die memory controller which will only benefit you in low latency areas such as gaming. And like i said, Gaming is NOT a CPU intensive sport. :)
 
I thought it was a bit odd to compare it to the 939 A64, b/c the 3500+ is only $346 vs. the Anandtech's quote of $850 for the 3.6 EM64T model. Hell, that very line is in the conclusion of the article, so I'm wondering why they even considered making the comparison in the first place.

Kind of like saying, oh wow, the Athlon XP loses to the FX-series. Kind of a no brainer?
 
Mav451 said:
I thought it was a bit odd to compare it to the 939 A64, b/c the 3500+ is only $346 vs. the Anandtech's quote of $850 for the 3.6 EM64T model. Hell, that very line is in the conclusion of the article, so I'm wondering why they even considered making the comparison in the first place.


Kind of like saying, oh wow, the Athlon XP loses to the FX-series. Kind of a no brainer?
Click to expand...

Because of the fact the 3.6 Ghz Prescott EM64T variant will perform in the same manner as the Xeon.

Anandtech said:
Although the Athlon 64 3500+ and the Xeon 3.6GHz EM64T processors were not necessarily designed to compete against each other, we found that comparing the two CPUs was more appropriate than anticipated, particularly in the light of Intel's newest move to bring EM64T to the Pentium 4 line. Once we obtain a sample of the Pentium 4 3.6F, we expect our benchmarks to produce very similar results to the 3.6 Xeon tested for this review.
Click to expand...
 
If I remember right, there was an article that was linked on the HardOCP homepage that had Intel stating that the EM64T chips won't be available for retail (only OEM servers).
 
Xrave said:
If I remember right, there was an article that was linked on the HardOCP homepage that had Intel stating that the EM64T chips won't be available for retail (only OEM servers).
Click to expand...

Yea that link was up like last week. The prescotts 64's are going to be servers and not desktops at least thats how it is right now.

I didnt read the article so I dont know how much the A64 lost to the Xeon (considering the Xeon IS a server cpu I would expect it to beat a desktop cpu in math instructions and what not, otherwise Intel is really screwing up, dont ya think BG?). However I think its kinda weird to conclude that the prescott 64 is gonna perform on par with the Xeon? What kind of marketing would that be? Hey all of you OEM system builders we have a $400 cpu that performs just as good as the $800 cpu. Hmm I wonder which one the OEM people are gonna buy? But like I said I didnt read the article so I have no idea what "ass handed" means but it much be pretty good considering the Xeon cost more than twice as much not to mention your probably (if your smart and care anyways) buy more expensive registered ram sticks :(

And oh Doom 3 is such a bad example of how cpu's make a difference in a game when the game is capped at 60fps (unless you took those numbers out of the [H]'s first article w/ the benchmarking where the game wasnt capped. I didnt care enough to go back and check you can correct me if im wrong...hell if wrong on any of these go ahead and correct me knowledge is power :) ) so after a certain cpu speed it really just doesnt matter what cpu you are using the same can kinda go for the gpu/vpu (depending on how high in reso and aa/af you play in) as far as capped games go.

However what you said about whats appropriate for making your decision on buying a cpu couldnt be more true, except for your comments on Intel being better at just about everythign cpu dependant. I dont have an A64 I have an AXP but let me tell you when im chilling with like 10-20 windows up and go switching from program to program I dont so my pc hiccuping the slightest bit. And yes those 10-20 windows have "intesive cpu tasks" going on (not all of them of course maybe 2-5?) not to mention I do have a oc'ed 2.4 C so I definitely now what im talking about as far as what I "feal" in that case (no bs/baisedness going on in that your open to your own opinion however if you have one, which im sure you do ;) ) and that P4 proc is running with ddr 500 btw.

Otherwise id say id agree with you on about two to three quarters of what you said. Not that that mite matter much to you anyways
 
burningrave101 said:
I just wanted to chime in and give you guys a little heads up on this whole A64 vs P4 gaming scenario in case you didn't realize it already.

The CPU has little to do with the performance in GPU intensive games, which the majority of games are, unless you are CPU limited.

In today's games like Far Cry, UT2004, DOOM 3, and so on, your NOT going to be CPU limited at higher resolutions with AA + AF enabled. You are GPU limited. At 1600x1200 or even 1280x1024 with high AA + AF, the P4's run just as fast as the A64's. The P4 rigs have even been beating the faster A64 setups in quite of few of the recent CPU scaling reviews. This could have to do with the simple fact that Intel has better chipsets and drivers.

If you will look at Hards DOOM 3 benchmarks you will see that the Athlon FX-53 with the 6800GT is only 1 fps faster then the 3.0c and 6800GT.

Gaming is a poor reference for a CPU buying descision unless you get your jollies off by sitting around playing at 800x600.

When you look at CPU performance you need to be worried about CPU intensive tasks. Not GPU intensive tasks like Gaming. There are only a handful of truly CPU intensive games out there such as Elder Scrolls III: Morrowind.

CPU intensive tasks include: Multitasking, Rendering, Compression, Encodeing, Productivity Software, Math Analysis, and so on. The Pentium 4 architecture is faster in the majority of these areas, especially with HyperThreading in multithreaded environments.

Anandtech just put up a new Article benching one of the new Xeon EM64T chips against an A64 3500+ in Linux 64-bit and the A64 litterally got its ass handed to it in math intensive applications.

http://www.anandtech.com/linux/showdoc.aspx?i=2158

The only real advantage of an A64 is the on-die memory controller which will only benefit you in low latency areas such as gaming. And like i said, Gaming is NOT a CPU intensive sport. :)
Click to expand...

I don't know where you are getting your reviews of the P4's beating the best AMD64's. Where is your links? Second of all, just about EVERY gaming benchmark I have seen the AMD64's OWN the P4's with ease. Oh and that BIASED review over at anandtech uses the SLOWEST 939 socket AMD cpu they have against intels best. The AMD cpu they used is $500.00 LESS. Why didn't they use the FX-53? Why didn't they do ANY gaming benchmarks? I'll tell you why, because if they would have compared it apples to apples the AMD would have moped the floor with the P4. Again. Not to mention the fact that the guy that did the review used OLD numbers for the 3500+. Even Inquirer Inside is upset over this biased review. http://www.inquirerinside.com/?article=17754

Fact of the matter is, the AMD64's OWN P4's in gaming anyday of the week. And since gaming is what I mostly use for a computer i'll be sticking to AMD and staying far away from the much slower P4's.

Quote from the Inquirer: If the author had said that the "fastest 3.6GHz Xeon trounced the slowest socket 939 Athlon 64 in math-intensive benchmarks at $500 more in cost", that would have put the paragraph in context. But as it currently stands, someone at Intel Corp. will no doubt use the AnandTech quote for the chip giant’s next round of Xeon promotional material. Intel will be happy, if it understands "happy".
 
burningrave101 said:
The CPU has little to do with the performance in GPU intensive games, which the majority of games are, unless you are CPU limited.
Click to expand...

For many GPUs, you're right.
But the CPU scaling article done here not long ago shows that with top end GPUs, you need a top end CPU.
If he's considering an X800/6800 (or even 9800/59xx) the CPU is very important to getting the most out of your card. Your gaming experiance may be very good with a mid range CPU, but the top end GPUs are most deffinately limited by those CPUs in current games.

http://www.anandtech.com/linux/showdoc.aspx?i=2158

The only real advantage of an A64 is the on-die memory controller which will only benefit you in low latency areas such as gaming. And like i said, Gaming is NOT a CPU intensive sport. :)
Click to expand...

(I'll ignore the fact that that review couldn't even copy the right results for A64 from previous Anandtech reviews in MySQL)

He asked about gaming, there are 0,gaming benchmarks in that review.
Look at any review, whether you compare PR to mhz, price to price, to best to best, Athlon64 is the better choice. Provided of course, he's looking at a high end GPU,
it he's throwing in a 9600 Pro, I agree, it doesn't matter, pick a 2.4C/A64 2800+, you won't notice the difference.
 
DemonDiablo said:
First of all everybody knows that the AXP line was supposed to go up against the P4 "A" and "B" procs not the P4 "C", "E", "EE" that your talking about. The A64 goes up against the P4 "C" and "E" while the AFX goes up against the P4 "EE" (hope your still with me ;) )

Secondly before you start to throw around dumbass comments about not knowing the fundatmentals about cpu arc. know that I am A+ cert. and that I am CSC major (thats all that you really need to know...for now) and that no shit the A64 has an integrtated memory controller :eek: So please dont try to lecture me about it nOObie

And where did you get this info about the P4 EE beating the P4 C and E by more than 15%? They used a P4 EE at 3.4Ghz but they didnt use any P4 "C" or "E" at 3.4Ghz? Just wondering what website you got those numbers from.
Click to expand...

Wow, you sure have a bad attitude.

I was referring to Anandtech's review, which is what the posts were about the in first place. It was that review which stated Doom 3 loves the cache.

In that review, they compared 3.2 Ghz Northwood, Prescott, and EE. There was a fair amount of performance separation amongst the three.

Oh BTW, you don't need to brag about your education. I have a EECS degree from a particular university on the East side of the SF bay. Unless your degree is from a particular university on the West side of the SF bay, you have nothing to proclaim superiority.

Degrees don't necessary mean you really understand what you're talking about. I once heard someone argue that most arcade games and other video game systems don't lock inputs to Vsync, using his education as his argument's main supporting point. Very naive, any well-designed embedded system that outputs video should lock inputs to Vsync (I hope you knew this already).

But honestly, your statement about L3 not being used is quite naive. Any sort of memory region, whether it be L2 or L3 cache, that's faster than dram will help the P4 and Athlon XP greatly.
 
superbooga said:
Oh BTW, you don't need to brag about your education. I have a EECS degree from a particular university on the East side of the SF bay. Unless your degree is from a particular university on the West side of the SF bay, you have nothing to proclaim superiority.
Click to expand...

Haha :D
 
The on-die memory controller kicks P4 ass. AMD64 is proven to be faster in floating point intensive apps whereas the P4 owns in productivity apps.
 
OMG ahahahah I just finished reading that AnandTech review. What a piece of shit. The Athlon school the P4 in the first round of testing then got it's ass handed to it. First of all who give a shit about synthetic benchmarks? Absolutely pointless if you ask me. And second, it's INCREDIBLY sad and shit-in-my-pants hilarious at the same time that Intel's SERVER hard-to-come-by chip hardly beats AMD's consumer proc :rolleyes: That review is a helluva lot like saying, "The fastest kenyian beat the average high school student at the mile."
 
All Intels are better even a P1 is better than the best amd 64 just becouse AMd is a crap company and I would never buy one. I would rather spend $2000 more on a 286 intel with math coprocessor than $1 for a AMD 3800+ or whatever the top ones is
 
I'm not tring to start any wars here but this is my oppinion in the AMD vs. Intel debate. CPU's rarely make a signifigant difference in gaming. My opinion is to buy the cheaper of the two and put that money toward a better GPU or more RAM. But I'm kind of a noob, but from my expiriance, this is what I have come to.
 
Phantum said:
OMG ahahahah I just finished reading that AnandTech review. What a piece of shit. The Athlon school the P4 in the first round of testing then got it's ass handed to it. First of all who give a shit about synthetic benchmarks? Absolutely pointless if you ask me. And second, it's INCREDIBLY sad and shit-in-my-pants hilarious at the same time that Intel's SERVER hard-to-come-by chip hardly beats AMD's consumer proc :rolleyes: That review is a helluva lot like saying, "The fastest kenyian beat the average high school student at the mile."
Click to expand...

Uh, not really. Xeons are nigh identical to their Pentium counterparts, the main differences being cache and multiprocessing capability.
 
Well i said it once and i guess i'm going to have to say it again lol. The CPU usually has little to do with gaming performance and an A64 isn't going to net alot better fps then a high-end Pentium 4 in gaming unless you play CPU intensive games which are rare. My favorite game Elder Scrolls III: Morrowind is a CPU intensive game but its one of the few and is probably one of the most CPU intensive games out there.

The reason people think A64's are so much better for gaming is because in reviews they turn the resolution down to 800x600 and turn the eye candy off. They do this so that the CPU is having to do alot more work to keep up with the GPU which is pumping out well over 100 fps. Now in reality who is going to play at these settings? Noone.

Now i want you to look at this CPU scaling guide that HardOCP just did here a little while back and compare the performance of the Pentium 3.0c to the A64 3500+ and then show me where the A64 is dominating the Pentium 4.

http://www.hardocp.com/article.html?art=NjMy

The Pentium 4 3.0c is able to compete with the A64 at higher resolutions with the eye candy enable because its the GPU that does the REAL work in games, not the CPU. Its good to have a fast CPU but you dont have to have the fastest to compete with the fastest.

A CPU's performance should be measured in CPU intensive tasks like i mentioned earlier. Encodeing and Rendering are two of the most CPU intensive tasks out there. The Pentium 4 is the king of Multitasking because of HyperThreading. The A64's gaming performance isn't coming from its processing power, its coming from its on-die low latency memory controller. Latency in games affects performance alot. Thats why its ALWAYS better to have low latency RAM over higher speed RAM with more bandwidth.

Now with all this said just let me finish by saying i'm planning to probably upgrade to an A64 next summer because i really like nVidia and their nForce motherboards and the new features they are coming out for them. I also like the fact that the A64's are starting to drop in price and they are 64-bit compatible for the future and they will give me better performance in my favorite game, Morrowind.

All in all, both processors have their strengths and weaknesses but i would say the Pentium 4 is still as good or a little better then the A64 because of being strong in the truly CPU intensive areas.
 
Well i chose a P4C when i built this computer because i like the little P4 sticker with HT and thought it would look cool wiht my case.

and i am not joking.. its the only difference that made me choose. since both systems performed about the same.. i figured... what the hell... i bounce back and forth all the time anyway, jsut comming off a AXP 2500+ system.

They are both fast, and whatever you get, you are not gonna notice.
 
You must log in or register to reply here.
Back
Top