P4 or AMD 64

[Thermalrock]

Originally posted by: CaiNaM
yes, you are.. and if you want to debate a point, why not come up with a logical argument? do the tests yourself. the proof, as they say, is in the pudding.

reading comprehension would also help. no one said the intel increases speed when performing a second task, so why pull that out of your ass to cloud the issue?

being a "smartass" because you have nothing valid to offer doesn't make you look "smart" in any way, shape, or form.

so im a smartass?. fine youre a dumbass. youre overrating hyperthreading so much its retarded. sure it helps some stuff but 20% in some special situations. btw you dont give proof for any of what you say. you claim hyperthreading does so much when it doesnt. why dont you link some tests that show youre right? you cant. the p4 is a single cpu. sometimes hyperthreading helps some and sometimes it even hurts the performance a lil. sure the athlon doesnt shine with two heavily p4 optimized processes running against a p4 running the same stuff. hyperthreading is not as good as you say and the other guys claims are even more ridiculous. ppl are asking something and will take what you type into account before making their purchase and youre giving them wrong facts. i just hope theyre not stupid enough to but their money where your mouth is.

 

[CaiNaM]

Originally posted by: Thermalrock

Originally posted by: CaiNaM
yes, you are.. and if you want to debate a point, why not come up with a logical argument? do the tests yourself. the proof, as they say, is in the pudding.

reading comprehension would also help. no one said the intel increases speed when performing a second task, so why pull that out of your ass to cloud the issue?

being a "smartass" because you have nothing valid to offer doesn't make you look "smart" in any way, shape, or form.

so im a smartass?. fine youre a dumbass. youre overrating hyperthreading so much its retarded. sure it helps some stuff but 20% in some special situations. btw you dont give proof for any of what you say. you claim hyperthreading does so much when it doesnt. why dont you link some tests that show youre right? you cant. the p4 is a single cpu. sometimes hyperthreading helps some and sometimes it even hurts the performance a lil. sure the athlon doesnt shine with two heavily p4 optimized processes running against a p4 running the same stuff. hyperthreading is not as good as you say and the other guys claims are even more ridiculous. ppl are asking something and will take what you type into account before making their purchase and youre giving them wrong facts. i just hope theyre not stupid enough to but their money where your mouth is.

dumbass? that's the best you can come back with?

you've done nothing bout wave the amd flag, pulled stuff out of your ass in an effort to back your bias point of view, but have not stated anything of substance to refute a single statement i've made.

i've been very specific about what i perceive as the strenghts and weaknesses of each cpu - even using acutal numbers such as amd often being 10-15% faster than intel in lower res gaming where the video card is not the bottleneck (instead of using terms that mean nothing other than show bias such as "smoke", "rule", etc.).

hell, the OP is even going with an amd platform in part because of my responses, and thanked me for my contribution. he can see both sides of the issue, why can't you? perhaps because your agenda is not to discuss what the OP asked or help with real information, but to push your amd bias on others?

if you want to refute or debate a point i've made, stick to the point and be specific; i'd be happy to discuss it. but if you're just going to rhetort with fanboy crap, then don't expect a cordial response. there's enough FUD and brand bias to go around already. we hardly need more n00bs to this board to come in and contribute to the idiocy.

 

[Thermalrock]

yea im hardly a noob and and you falling for the p4 hyperthreading tv commercials isnt my fault either. show me one single unbiased site that says hyperthreading does help a lot. you cant. run around recommend a cpu that doesnt cost less and needs twice the power to perform worse overall and uses a socket that prolly wont live half as long. the a64 is the better choice. for every p4 optimized software theres another one that gets the job done as fast or faster on an athlon. im not waving the amd flag either. early last year i would have said northwood over xp, before that i would have recommended a palamino over a willamette. i would tell ppl to get a pentium m if they plan on getting a notebook. while the p4 might have slight advantages in few situations those are mostly artificial because the software was optimized for the p4. any pre p4 code runs better on an athlon. any software thats not optimized for a specific cpu runs better on an athlon. the stuff that is heavily optimized for p4s runs faster ona p4 but the p4 beats the athlon by a rather surprisingly small margin. hyperthreading isnt as good as you say nomatter what and you cant prove this isnt true it thats why you talk around say do your own tests and stuff when i actually see p4hts all the time in the cip pools of my university. im just not gonna try to make the difference bewteen two cpus seem smaller than it is for odd reasons. i just like whatever performs better and right now thats the athlon. as i said 18 months ago it was the p4. not saying you cant do anythingy ou want on a p4 but the athlon is the better choice right now and will be till intel will come up withs omething thats actually better. and you bet as soon as they do i will recomment their product if amd doesnt have anything that performs better by then. show me how hyperthreading is this big of a deal. one unbiased benchmark that proves it. there are none and you know it else you would have. it helps sometimes a lil and sometimes its even alil counterproductive to have it enabled. athlon got awefully close in encoding decoding stuff and where he already had the lead he made it bigger. all that with using only have the energy which is good for your electricity bills and this planet too btw. and youre not sacrificing speed at all. and even intel knows netburst is a dead end. 4ghz canceled, tejas canceled. now show some prove or shut up. complaining about me not linking any evidence when you havent is ridiculous. show me an unbiased site that says ht is as good as you make it out to be and ill go look at it.

 

[Kovie]

Paragraphs are your friend. 🙂

 

[CaiNaM]

> yea im hardly a noob and and you falling for the p4 hyperthreading tv commercials isnt my fault either

umm.. there are no "p4 hyperthreading tv commercials"

> show me one single unbiased site that says hyperthreading does help a lot. you cant

first, the burden of proof is yours. i made an observation based on owning and using both daily. you imply i'm lying; you prove it. you can't.

second, your statement speaks for itself. any "site" that says it helps is obviously "biased" in your opinion based on your own words.

at any rate, here's one which someone else states was either biased or ill conceived, however they never offered proof of either:

Several interesting things about this chart jump out after a little study. First, both the top and the bottom data lines are both from runs with the Athlon 64. When WME9 was running, the Athlon 64 averaged less than 4 frames per second. We did see one large spike in frame rate, but the curve pretty much remained under 4 fps for the majority of the run. All three Pentium 4 processors performed more poorly when running Flight Sim 2004 solo, but managed to average around 17 frames per second while WME9 was chugging along in the background. The other interesting data point is that Prescott's average frame rate of 17.2 fps when multitasking was essentially the same as the 3.2GHz P4EE's 17.5 fps. Of course, the frame rate dipped on occasion, but the point here is that Hyper-Threading clearly has a major impact. link

kyle over at HardOCP had similar comments:

Originally Posted by Kyle
Multitasking

Every time I write about how Intel Pentium 4 CPUs with HyperThreading dominate in multitasking, I get mad email from AMD fans telling me otherwise. I have had a Pentium 4 CPU in my main work system, which is used for everything I do including gaming, for a couple of years till back in July. I put in an Athlon 64 FX-53 to compare real-world experiences. The difference in how multitasking is handled is much like night and day. Intel?s HyperThreading makes easy work of multitasking where the Athlon 64 simply flounders. I was used to encoding movies and music while I went about my normal work tasks and the Athlon FX system was having none of that. I found many places where the system would ?chug? due to the CPU handling one task while I had another one I wanted to focus on. Sure, you can go in and change the way the CPU handles the particular tasks you have open, but that is simply a pain to get done every time you want to use certain combinations of software.

Now with all that said, I have gotten very used to the way the Athlon 64 FX handles my daily computing needs and to some extent I have formed my needs around its performance. Do I still want my Pentium 4 with HyperThreading back? Absolutely. I think the Pentium 4 with HyperThreading is the best solution for those people that work with many different applications open at once that can be CPU intensive even when working in the background.

i'm kind of like kyle on this matter - i've gotten used to what i consider it's weakness, and to some extent i've also changed the way i do things around it's behavior. i also have an HT p4 sitting right next to my athlon which alleviates this issue. not everyone has that luxury however.

here's a good example i just ran across in a forum post:

Hyper Threading and System Loading:

Generally speaking, yes, in a multi-tasking environment that isn't I/O restricted (meaning you aren't doing something that slams the hell out of the hard drive), the P4 is going to out-perform the Athlon 64 (or will close the gap considerably if the Athlon 64 outperforms it in these functions when evaluated one on one.

Folding @ Home is actually a fabulous case performance for Hyper Threading and it doesn't require the program to be set in real time in order to see this. Here is what happens:

Take two FAH threads. Comparisons will look like this:

Athlon 64:
Time to Finish a Frame of Thread 1: 2 min, 30 sec.
Time to Finish a Frame of Thread 2: 2 min, 30 sec.
Time to finish a frame of T1 & T2 Simultaneously: 5 minutes

(I am simplifying numbers to make my point).

P4 w/ HT
Time to Finish a Frame of Thread 1: 3 minutes.
Time to Finish a Frame of Thread 2: 3 minutes.
Time to Finish a Frame of T1 & T2 Simultaneously: 4 min, 45 seconds.

The point here is that while the Athlon 64 outperforms the P4 on both Frame 1 AND Frame 2 when evaluated separately, the P4 (thanks to HT) will outperform the Athlon 64 by a small margin when running them together.

Hyper Threading is not the be-all, end-all of performance that Intel's marketing makes it out to be at times, and there are many scenarios where it offers little to no benefit. There are also significant scenarios (DVD encoding, some multi-tasking, 3D rendering) where it offers significant benefit.

again, i agree.. it's not a "be-all, end-all" - i never said it was, but in certain circumstances it is indeed quite beneficial when compared to how the athlon64 handles dual threads.

For the multitasking scenario, we chose to run Norton AntiVirus in the background while using Windows Media Encoder 9 in the foreground to convert a 30-second AVI clip to a high-quality WMV file. We report the time it took to run just the video encode by itself and with NAV running in the background. The results show that Intel's Hyper-Threading clearly pays off. The Pentium 4 took about a minute less time to run the multitasking test than the Athlon 64 FX-51 systems did. Copyright © 2004 Ziff Davis Media Inc. All Rights Reserved. Originally appearing in PC Magazine.

there's but a few statements/opinions offered similar to mine. you have offered NONE; not even your own observations, only your own biased opinions based on.. i dunno.. you've offered nothing. 'nuff said.

i could continue, but the rest of your post offers nothing more than the same regurgitaged opinionated crap which is hardly worth commenting on.

 

[Thermalrock]

as i said i use p4s too. youre quoting guys who own a p4......... great theyre gonna say ok i bought a cpu for the same price that has no advantage. and the other one lololol:

'Time to Finish a Frame of Thread 1: 2 min, 30 sec.
Time to Finish a Frame of Thread 2: 2 min, 30 sec.
Time to finish a frame of T1 & T2 Simultaneously: 5 minutes

(I am simplifying numbers to make my point).

P4 w/ HT
Time to Finish a Frame of Thread 1: 3 minutes.
Time to Finish a Frame of Thread 2: 3 minutes.
Time to Finish a Frame of T1 & T2 Simultaneously: 4 min, 45 seconds. '

are you just stupid and dont see that this UNDERLINES exactly what i said? the ht does it ~20 per cent faster than it would with out ht EXACTLY what i said. that it still doesnt manage to be 20% faster than the athlon btw. god youre dumb.

 

[carlosd]

Originally posted by: CaiNaM

Originally posted by: Thermalrock
pretending the p4 ht is a dual cpu system again....

nope, just stating facts. try it, then you won't have to fight ignorance 🙂

unless of course you're one of those who only acknowledge fact when it fits neatly into what they want to believe.

the athlon is a great chip, but to claim it has no weakness is a fallacy. it may not affect a great number of users who simply never use their hardware in that way (i mean lets face it, the majority of users don't do much more than browse the net, exchange emails, and play the occasional game), but there are those that demand more. if that weren't the case, there'd be no reason for amd to pursue dual core designs - they'd simply ramp up clock speeds. but even tho many 'amd enthusiasts' prefer to pretend it's not there, amd knows it's a weakness, and dual core will likely eliminate it.

If HT will work that way intel wouldn't have taken the dual core path, don't you think so. Regarding HT its not a dual core CPU and won't perform nearly the same.

 

[CaiNaM]

Originally posted by: carlosdIf HT will work that way intel wouldn't have taken the dual core path, don't you think so. Regarding HT its not a dual core CPU and won't perform nearly the same.

i certainly never compared HT to dual core, so am not sure what your point is. i would venture to say logic dictates a dual core cpu is > one physical core plus logical core, tho we'll have to wait for them to actually appear before we can do hard comparisons on the differences.

 

[carlosd]

Originally posted by: CaiNaM
> yea im hardly a noob and and you falling for the p4 hyperthreading tv commercials isnt my fault either


here's a good example i just ran across in a forum post:

Hyper Threading and System Loading:

Generally speaking, yes, in a multi-tasking environment that isn't I/O restricted (meaning you aren't doing something that slams the hell out of the hard drive), the P4 is going to out-perform the Athlon 64 (or will close the gap considerably if the Athlon 64 outperforms it in these functions when evaluated one on one.

Folding @ Home is actually a fabulous case performance for Hyper Threading and it doesn't require the program to be set in real time in order to see this. Here is what happens:

Take two FAH threads. Comparisons will look like this:

Athlon 64:
Time to Finish a Frame of Thread 1: 2 min, 30 sec.
Time to Finish a Frame of Thread 2: 2 min, 30 sec.
Time to finish a frame of T1 & T2 Simultaneously: 5 minutes

(I am simplifying numbers to make my point).

P4 w/ HT
Time to Finish a Frame of Thread 1: 3 minutes.
Time to Finish a Frame of Thread 2: 3 minutes.
Time to Finish a Frame of T1 & T2 Simultaneously: 4 min, 45 seconds.

The point here is that while the Athlon 64 outperforms the P4 on both Frame 1 AND Frame 2 when evaluated separately, the P4 (thanks to HT) will outperform the Athlon 64 by a small margin when running them together.

You call a difference of 15 seconds outperforming?? this is a 5% difference please!!

 

[CaiNaM]

Originally posted by: Thermalrock
as i said i use p4s too. youre quoting guys who own a p4......... great theyre gonna say ok i bought a cpu for the same price that has no advantage. and the other one lololol:

'Time to Finish a Frame of Thread 1: 2 min, 30 sec.
Time to Finish a Frame of Thread 2: 2 min, 30 sec.
Time to finish a frame of T1 & T2 Simultaneously: 5 minutes

(I am simplifying numbers to make my point).

P4 w/ HT
Time to Finish a Frame of Thread 1: 3 minutes.
Time to Finish a Frame of Thread 2: 3 minutes.
Time to Finish a Frame of T1 & T2 Simultaneously: 4 min, 45 seconds. '

are you just stupid and dont see that this UNDERLINES exactly what i said? the ht does it ~20 per cent faster than it would with out ht EXACTLY what i said. that it still doesnt manage to be 20% faster than the athlon btw. god youre dumb.

hahaha.. you're lack of logic is only exceeded by your ability to change your arguments and pull stuff out of your ass. not to mention all the other relevant points which continue to disregard...

where did i say intel was 20% faster running F@H?

where did you say HT was a benefit? i assumed your "like if you encode something and it would take an hour and you wanna speed it up all you gotta do is start doom3.." your comment was sarcasm, as the alternative would be you were serious, contradicting all your other posts - i would never have imagined it possible that anyone could be THAT clueless... i guess i was wrong :Q

 

[Thermalrock]

oh and cainam before you post bs again all you gotta do is look at the anands benchmarks. anyway im done arguing with you go buy a p4 i couldnt care less, ill buy an intel when they come up with something better than amd. and the statements you copied from ppl in this forum, .... why are you so ridiculous? there are way more pro amd posts in this forum and you know that, even in this very thread and you complain about me not copying them for you? and you wanna know why there are more amd posts? because amd has the better cpus at the moment. 18 months ago intel had the lead and maybe theyll get it back in 2005 who knows. go look at benchmarks and dont read so many posts of ppl who just bought an intel and take them as facts.

 

[CraigRT]

Originally posted by: coinz
alright thank you..yeah his budget allows the 3500+ he won't be doing any overclocking so I think I got my stuff to finally put together this system

'grats

 

[Thermalrock]

Originally posted by: CaiNaM


hahaha.. you're lack of logic is only exceeded by your ability to change your arguments and pull stuff out of your ass.

where did i say intel was 20% faster running F@H?

where did you say HT was a benefit? i assumed your "like if you encode something and it would take an hour and you wanna speed it up all you gotta do is start doom3.." your comment was sarcasm, as the alternative would be you were serious, contradicting all your other posts - i would never have imagined it possible that anyone could be THAT clueless... i guess i was wrong :Q

ok youre just retarded. im sure now. i said that i said, go reread dumbass, that ht is llike 20% faster in some cases.... and you post numbers that prove what i say. anyway learn how to read. youre embarrassing yourself. each post you make lets you look more stupid than the one before.

 

[CaiNaM]

Originally posted by: Thermalrock
oh and cainam before you post bs again all you gotta do is look at the anands benchmarks.

why? they don't deal with the point of our discussion, tho they do back up my statements saying comparably rated amd64s are 10-15% faster than intel.. still, you're again trying to change the subject matter in order to support your bias.

anyway im done arguing with you go buy a p4 i couldnt care less, ill buy an intel when they come up with something better than amd. and the statements you copied from ppl in this forum

own anything you want, that's your right. i only object to your using bias comments and FUD to coerce others who are seeking valid opinions.

why are you so ridiculous?

if "ridiculous" is offering relevant, unbiased, factual information.. you'd have a point.

there are way more pro amd posts in this forum and you know that, even in this very thread and you complain about me not copying them for you?

nope, just offer factual observations, that's all. i never asked for links to opinions, only for anything factual to support your contention by conclusions are false. you have not, so i can only conclude you cannot.

and you wanna know why there are more amd posts?

no

because amd has the better cpus at the moment. 18 months ago intel had the lead and maybe theyll get it back in 2005 who knows.

did i ever say intel had better cpus? or that amd's cpus were worse? nope. once again you continue to ignore the actual content of what i've stated.

go look at benchmarks

you're going in circles, once again repeating things which have no relevance to the subject. see my reply to the first time you asked this.

and dont read so many posts of ppl who just bought an intel and take them as facts.

my initial responses were based on my owning and using both platforms daily. while you failed to offer anything supporting your opinion, you asked for me to do so. while the burden of proof is yours (and you've failed miserably, having made a half dozen posts without ever being able to support anything you've said), i simply offered both reviews and opinions of others which support my experience, just as you asked.

ok youre just retarded. im sure now. i said that i said, go reread dumbass, that ht is llike 20% faster in some cases.... and you post numbers that prove what i say. anyway learn how to read. youre embarrassing yourself. each post you make lets you look more stupid than the one before.

lol.. the best you can do? you should give up.

 

[Thermalrock]

you dont give unbiased opinions you copied crap intel owners posted on this board and call them unbiased. you cant give one link that proves anything wrong ive said. and you cant. go copy a few more posts of ppl who write about how their intel can multitask and compare it to dual cores. the funny thing is that you dont realize how pathetic you are. you even included a post that says exactly what i said and didnt get it. then you replied something retarded proving youre either too stupid to read or comprehend...
im not gonna reply to your multitasking bs/hyperthreading bs/ quoting kids with a p4 on a message board when asked for credible sources stating anything to your pathetic claims anymore. youre a waste of time. you can take this post apart line for line again and reply silly crap noone cares, hyperthreading will still not do any more than up to 20% in some special situations and you still wont be smart enough to comprehend that youre calling me biased for saying the exact same thing someone you quoted to 'prove your point' said. just this time you call it an 'unbiased' opinion, the exact same thing i said. and yes you are ridiculous. just not for quoting unbiased opinions which you btw havent done yet with the exception of that one but ffs you still dont get that it just underlines my statement. ~~~~~~~~
way to go dumbass.

ps i repeat myself frequently cuz you proved to be pretty slow and you still dont get it so obviously i havent repeated myself enough. hyperthreading still doesnt do more than i said it does and you naturally didnt come up with any proof that what i said was biased or false since it wasnt. youll most likely still continue typing crap and quoting crap presenting it as facts but it wont change anything. you even forgot who started this and said i would have to prove myself when i made a post unrelated to you and you jumped on me calling me biased. it would be on you to back up that claim up but you cant cuz i didnt say anything other than that an athlon would be a better choice, and that hyperthreading does not help as much as some ppl here make it out to do. anyway im done with you nothing you say is worth anything.

 

[uOpt]

Originally posted by: Thermalrock
i take it all back and im sorry for being a non believer, a p4 because of hyperthreading can of course run 2 threads at once without losing any speed. ht doesnt help about 20% like i said now actually if you start a second process the p4 wakes up and performs each one faster than he would only one. like if you encode something and it would take an hour and you wanna speed it up all you gotta do is start doom3 and the encoding will take only an hour while the game play will run at a few more frames............

some of you guys are pathetic.

Here are the numbers you need:
http://forums.anandtech.com/me...8&threadid=1426699

Basically hyperthreading is very limited when it comes to two CPU hogs but when running a game in the foreground and a CPU hog in the background it really shines.

It is likely, but hard to measure, that intense 2G graphics in the foreground show similar results.

 

[uOpt]

BTW, if you want fast encoding for the money, get a Celeron D.

 

[Thermalrock]

what? youre suggesting what exatcly with those numbers? double the frames? what did running the game do to the speed of the background application? those numbers seem fishy on first sight. would you mind giving a short conclusion as i dont feel like reading that entire thread? ht makes things faster by how many per cent according to you?

 

[Artanis]

can you ever stop the useless fight?

 

[uOpt]

Originally posted by: Thermalrock
what? youre suggesting what exatcly with those numbers? double the frames? what did running the game do to the speed of the background application? those numbers seem fishy on first sight. would you mind giving a short conclusion as i dont feel like reading that entire thread? ht makes things faster by how many per cent according to you?

Those fishy numbers are what I measured myself having both the P4 and the AMD64 in my home. I'd say they are a lot more reliable than some of the speculation around here 😉

You can extract the framerate numbers easily. I don't feel like summarizing as long as you seem to imply I claim a framerate raise.