Sign in with
Sign up | Sign in
Your question

Attacking Intel

Last response: in CPUs
Share
Anonymous
a b à CPUs
March 27, 2002 3:11:53 PM

I am sick of the peoples who attack Intel for no reason. IPC do not matter, it I am performance of the end. Pentium 4 has the abundance of that one. It also has IHS, the thermal protection, and SSE2, three things that AMD does not have (still). FatBurger put a challenge ahead to prove that Athlon has not the performance best, and same that many trolls attack Intel as being a bad processor, they do not have the courage to accept the challenge. It is of my opinion that is one hypocrite and attacking Intel for no reason.

P.s. Sorry for deficient mine English, I am of Portugal. I read Tom's Hardware and these forums for some hour, but I did not write here because mine English is not good.

More about : attacking intel

March 27, 2002 3:21:01 PM

So whois attacking intel for no reason that's not involved in an post started by an intel_troll ? And i couldn't disagree with you more =).

<font color=red>:</font color=red> <font color=white>:</font color=white> <font color=blue>:</font color=blue>
March 27, 2002 3:21:40 PM

Why would IPC not matter? Parallelism has always been a factor in CPUs. If you think it doesn’t work stick your head in a beehive and ask yourself how much honey can one bee produce? I don’t hate Intel but I do hate disinformation. In the end you’ll have to make your own decision on what processor you like. If you ignore half the equation you will be making a half assed decision.

All errors are undocumented features waiting to be discovered.
Related resources
March 27, 2002 3:36:44 PM

loooooooooool @ IPC do not matter, then what is matter? clock speed? if so then how you prove that AMD performs better than Intel?

how many computer users arround the world getting the benefit of the SSE2?

Intel is not a bad processor, but when they say they are the best and have fastest pocessors so we get bothered from that because this is not true

regarding fat burger challenge and overclocking, unfortunaly most of the users here are talking about overclocking and they forget that only 5% or less of users overclock their machines and this apply only for single anf home users, i never heard a company who own 10's of computers is thinking to overclock or they know what is overclocking. So if we get this fact that overclocking is limited to a specific type of users and specific numbers of them then AMD is the leader.

AMD is not perfect in every application but in most it leads, unless the application has a special code for Intel.

again, the way intel uses to advertise for its processors and the way they go make the some people feel bad

wish if there was UnDo in the life
March 27, 2002 4:12:13 PM

Quote:
regarding fat burger challenge and overclocking, unfortunaly most of the users here are talking about overclocking and they forget that only 5% or less of users

However, I do feel that at least most of the people here are willing to overclock, which was why I felt comfortable presenting that challenge. If this was a forum at Cnet or something similar, I wouldn't have.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 27, 2002 4:27:31 PM

Quote:
However, I do feel that at least most of the people here are willing to overclock, which was why I felt comfortable presenting that challenge. If this was a forum at Cnet or something similar, I wouldn't have


i do not have any objection about overclocking and who i am to determine and talk about overclocking, but what i wanted to say that AMD XP processors is better than Intel P4 in terms of performance if you are not overclocking and this is at least for the time being and we do not know what the situation will be with tbred.

and since most of the people around the world do not know or think about overclocking then you can not rate the cpu performance when it is overclocked

wish if there was UnDo in the life
March 27, 2002 4:36:42 PM

Quote:
but what i wanted to say that AMD XP processors is better than Intel P4 in terms of performance if you are not overclocking


If it was impossible to overclock, I would never have even considered buying a P4.
I feel that since "most of the people around the world" don't read this forum or understand half of what is said in here anyway, I can make my challenge safely.

I'm not disagreeing with you, just clarifying.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 27, 2002 4:39:00 PM

yeah I agree, it's performace in the end that counts. However amd zealots enjoy pulling the card of "p4's ipc sucks compared to athlon's" when the p4 pulls ahead.

I think it's a good time for the cpu market though, we have 2 processors which are very close in performance. Both companies are trying to outdo each other. This can be good or bad. The good part is it drives prices down, and higher clocked cpu's are released more quickly. The bad part is we get shady actions from both companies, such as intel saying that mhz is everything or amd creating the confusing pr rating.

<i>My life wasn't complete untill I tried sse-2 optimized pong</i>
March 27, 2002 4:57:10 PM

I think Intel makes great CPU. The only reason I build my system with an AMD CPU is that I do not want to pay a few hundred dollars more for an similar performance system with Intel Processor.
March 27, 2002 5:01:57 PM

Quote:
FatBurger put a challenge ahead to prove that Athlon has not the performance best, and same that many trolls attack Intel as being a bad processor, they do not have the courage to accept the challenge.

Correct me if i'm wrong,but that chalange is for the Tbred,no?

I have yet to see one and I work for Tier 1 AMD disty...



The faster a computer is, the faster it will reach a crashed state :eek: 
March 27, 2002 5:10:46 PM

Quote:
However amd zealots enjoy pulling the card of "p4's ipc sucks compared to athlon's" when the p4 pulls ahead.

I don’t think anyone who replied to this post stated anything of the sort. Some did state the IPC matters. Its just like size, you know what I’m talking about, identical to the misrepresentation of length to width when volume is considered.


All errors are undocumented features waiting to be discovered.
March 27, 2002 5:14:55 PM

Quote:
I do not want to pay a few hundred dollars more for an similar performance system with Intel Processor.


And you don't have to, check xxsk8er101xx's thread for pricing on similar systems. Intel is still more expensive, but a "few hundred dollars" is a bit exaggerated.

Quote:
Correct me if i'm wrong,but that chalange is for the Tbred,no?


One was, the other was for any Athlon to out-perform my OCed 1.6A.

BTW, who do you work for?

Quote:
I don’t think anyone who replied to this post stated anything of the sort.


No, but it has been said many times before.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 27, 2002 5:27:25 PM

Quote:
BTW, who do you work for?

If this is directed at me, I have to dance around the subject. Does work not imply a job or can one just work for the enjoyment of the task? Should one measure ones salary in the Dot Com boom to ones salary in the Dot Com bust? I work for the man who cares the most about me.


All errors are undocumented features waiting to be discovered.
March 27, 2002 5:32:33 PM

No, it was directed to Varlo. No dancing required. At least not from you. :wink:

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 27, 2002 5:55:18 PM

I have nothing against Intel processors its just thier PR tactics and marketing crap they spew out concerning Mhz as the only viable metric.

I'd sock it to Intel anyday..they need a slap round the face...who are they are kidding ?

<font color=purple>~* K6-2 @ 333MHz *~
I don't need a 'Gigahertz' chip to surf the web just yet ;-)</font color=purple>
March 27, 2002 6:24:47 PM

You think a company is not going to promote thier product? Are they supposed to say thier product is crap? I would most companys exagerate at least a little on what thier products do.
March 27, 2002 6:57:19 PM

Also, most of the people around the world do not really care what "IPC" actually means. They do not want to "unlock" their Athlon XP multiplier for overclocking and it was Tom's Hardware that published the related information to do that.

Searching for the true, the beautiful, and the eternal
March 27, 2002 7:12:57 PM

At least you CAN unlock it, Intel won't let you at all unless you get them to send you a specially hand-picked chip.
So with AMD you have more freedom in overclocking enthusiasm.

--
For the first time, Hookers are hooked on Phonics!!
March 27, 2002 7:25:19 PM

I never tried to unlock my Athlon XP 1600+ because the warranty clearly states it voids everything and I paid nearly $200 for it.

It was easy for me to overclock a Pentium 120MHz to 200MHz and Celeron 300A to 504MHz. Many people around me run Pentium 4 1.6a at 2.4GHz. I never buy an Intel CPU for more than $150 - I mean not yet.

Searching for the true, the beautiful, and the eternal
March 27, 2002 10:49:21 PM

Quote:
regarding fat burger challenge and overclocking, unfortunaly most of the users here are talking about overclocking and they forget that only 5% or less of users



--------------------------------------------------------------------------------

However, I do feel that at least most of the people here are willing to overclock, which was why I felt comfortable presenting that challenge. If this was a forum at Cnet or something similar, I wouldn't have.



No one can even take the fb challenge till the tbred is released lol.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
March 27, 2002 10:50:50 PM

Quote:
And you don't have to, check xxsk8er101xx's thread for pricing on similar systems. Intel is still more expensive, but a "few hundred dollars" is a bit exaggerated.


Its all in the motherboard, I posted a combo which was better performing than the p4 combo for 200 less, so again, it depends on the motherboard.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
March 27, 2002 10:52:32 PM

Quote:
I never tried to unlock my Athlon XP 1600+ because the warranty clearly states it voids everything and I paid nearly $200 for it.

It was easy for me to overclock a Pentium 120MHz to 200MHz and Celeron 300A to 504MHz. Many people around me run Pentium 4 1.6a at 2.4GHz. I never buy an Intel CPU for more than $150 - I mean not yet.



Intels warranty CLEARLY states that overclocking it VOIDS the warranty, just like amds, if you overclock the fsb yoy VOID your warranty.

You can LIE to intel/amd if you return your chip, but you still violated your warranty. Comprehende!

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
March 28, 2002 1:24:55 AM

Again you are drifting the subject here...
Absolutly remotely nobody here buys AMDs for their warranties, JUST ASK. We don't care for them because we know what we do is on our responsibility. Otherwise the CPU forum would be dull and inexistent, just a forum to recommend new CPUs, nothing else like enthusiast discussions. So whether you would think AMD has anything wrong in their overclocking, it applies as well to Intel when doing so, and thus the unlocking of AMD CPUs is bigger better plus for AMD. I can just wait to see how Matisaro will enjoy unlocking and upping the FSB+Multi in sync with RAM, taking on Fatty's challenge without having to get an insane bus totally asynchronised.

--
For the first time, Hookers are hooked on Phonics!!
March 28, 2002 1:55:32 AM

Guerreiro...the performance of a CPU is determined by IPCxFrequency. The Athlon executes NINE instructions every clock tick, the Pentium 4 executes SIX. IPC is half the equation.

What's <A HREF="http://www.vanshardware.com/" target="_new">this</A> all about?
March 28, 2002 1:59:04 AM

Of course if they can make the instruction shorter and thus use less IPC per clock to fully pass it, a P4 can do as well. But if apps were properly coded for Athlon's IPC, it would be significantly above the P4 for a while, because it carries short high-definition instructions, all in 9 per clock. Again any could be better than other, but for now, Athlon's IPC is substantially better. No one will argue over FPU for Athlons now will you?

--
For the first time, Hookers are hooked on Phonics!!
March 28, 2002 2:46:31 AM

One challenge can be done, one cannot.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 28, 2002 3:09:35 AM

Oh yeah, I forgot you have 2 challenges, my bad.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
March 28, 2002 3:16:54 AM

That's ok, you're definitely not the only one.

The Thoroughbred one is the more interesting of the two anyway.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 28, 2002 7:42:21 AM

Sorry for being one of that nobodies.

Searching for the true, the beautiful, and the eternal
March 28, 2002 9:01:49 AM

Quote:
Also, most of the people around the world do not really care what "IPC" actually means. They do not want to "unlock" their Athlon XP multiplier for overclocking and it was Tom's Hardware that published the related information to do that.


agree for (the people around the world do not really care what "IPC" actually means) and it is the same for not knowing about overclocking, but this is what i wanted to say that most people ask which one performs better so it is AMD which perforsms better regardless of overclocing.

if you want to involve the buyer in overclocking and sse2 and other stuff then you have to tell him about quanti speed and IPC and the other stuff

wish if there was UnDo in the life
March 28, 2002 9:22:30 AM

Next time I see an AMD CPU with T-bred core maybe. Not now.

Searching for the true, the beautiful, and the eternal
Anonymous
a b à CPUs
March 28, 2002 2:25:36 PM

Quote:
Guerreiro...the performance of a CPU is determined by IPCxFrequency. The Athlon executes NINE instructions every clock tick, the Pentium 4 executes SIX. IPC is half the equation.

I agree to what you says, but you he did not prove that its point that Pentium 4 is not good AMD is better in an area, and Intel is better in another one. The performance and the value of the end are what it matters, and AMD is not the only company that has that one.
Anonymous
a b à CPUs
March 28, 2002 2:28:03 PM

Quote:
I think Intel makes great CPU. The only reason I build my system with an AMD CPU is that I do not want to pay a few hundred dollars more for an similar performance system with Intel Processor.


It does not have no reason to pay much more for a processor of Intel. Only that he is extremely more expensive is upper, the 2.2 or soon the 2.4.
Anonymous
a b à CPUs
March 28, 2002 2:30:42 PM

Quote:
loooooooooool @ IPC do not matter, then what is matter? clock speed? if so then how you prove that AMD performs better than Intel?


Ipc and clock speed both matter, I did not say in another way.

Quote:
how many computer users arround the world getting the benefit of the SSE2?


Much many, please look at a stack of software using SSE2 and you it will see.
Anonymous
a b à CPUs
March 28, 2002 2:32:33 PM

Quote:
So whois attacking intel for no reason that's not involved in an post started by an intel_troll ? And i couldn't disagree with you more =).


Why you calls me one troll? I am not one troll, me I am only trying to make peoples to think clearly on the facts.
March 28, 2002 6:00:39 PM

Quote:
Much many, please look at a stack of software using SSE2 and you it will see

That is at its best laughable. There simply isn't 'enough' to say SSE 2 is worldwide adopted. I could care less to buy a processor to make Black And White run with better performance and image quality (5% better), or having Tribes 2 run smoother when any card of today's generation can run it without the CPU's help. The other SSE2 apps are generally programming apps and rendering, but even then, the only way you can say it's used, and is optimizing the P4, is to check if it does so per clock. I had a statement done about a week ago in a topic, saying about how people should not go ranting apps are optimized for P4 unless they prove so per clock. Forget it, SSE2 is definitly not as successful as expected, and part of it comes from the fact the architecture of the P4 currently is stalling it from being better.

--
For the first time, Hookers are hooked on Phonics!!
Anonymous
a b à CPUs
March 29, 2002 5:18:09 PM

Quote:
I had a statement done about a week ago in a topic, saying about how people should not go ranting apps are optimized for P4 unless they prove so per clock


Please explain why optimization is important to "per clock". I of not understand this.
March 29, 2002 5:32:24 PM

Alright...okay...try to imagine this.

You have two race cars custom build against each other. One has a V6 engine while the other sports a V8. Now before the race, the maker of the car with the V6 can advertise that it has higher "top speed", but takes more time to throttle up. However, in the shorter run, the V8 has a better speed/acceleration ratio. Now in the race, the V8 undoubtedly would've beaten the V6.

Lets compare that to CPUs now. IPC is a term meaning "instructions per clock." A clock is a very short period of time when the CPU is processing instructions. Now this perios is extremely small, I believe only 1 nanosecond, or 1,000,000,000,000,000th of a second. Per clock, the Athlon has more instructions completed than the PentiumIV. The arguement over this matter is that the Athlon has a higher IPC but lower overall Mhz, where are the PIV supports a higher Mhz, but far lower IPC. CPU optmization is the task of programmers making a program able to utilize the CPU special instructions from SSE and SSE2, allowing the program to run faster in turn. Per clock IPC is extremely important for determining CPU utilization in conclusion.

"When there's a will, there's a way."
March 29, 2002 5:52:48 PM

Ok here is the copy paste of my previous rant:

Quote:

Another thing is, even if apps are seemlessly "optimized" for the P4, or SSE2 apps, I will not beleive something is optimized until it actually does more at the same clock speed. Yes I know I should not bring it up, but the fact and the only way to see something optimized, it means when it does much more at the same level. So if P4-optimized apps run better per clock on AthlonXP, then these apps are regretably NOT optimized at all for P4. Same applies for SSE2 and the rest.
EX: Take 2 big fields of corn (X,Y). Both grow corn very well, and although the X bigger one seemlessly does more, it's bigger so we tend to think it does more per chunk or acre. That's how the P4 is, and we tend to think the app using it is optimized for it. Now cut X corn field to the same size as Y, which is growing at the same rate as the X one, we then see that both are the same. Now let's say the smaller Y corn field was equalling the corn grown on the X one (when X was big), THEN we know that this corn field has premium ground, and that it is the optimum field to use and to continue increasing its field size than X's. Unless corn field X when it was the big one, was actually optimized for real and that when cut down to Y size, turned out to render more corn, then this one is optimized, and then we know that this app is optimized for P4.

Again I know clock per clock should not be discussed anymore, but if we want to discuss optimized apps, we have to do so to prove that they are optimized, which is why Ray says that some things run better on the 2GHZ model because some apps are using the P4's instruction set, but if I don't see proof per clock, then effectively the AthlonXP 2000 is still the better choice.

There may be some things I said that were off the topic here, so just focus on the optimization example, not what I said of other people!

--
For the first time, Hookers are hooked on Phonics!!
March 29, 2002 5:59:08 PM

Quote:
Per clock IPC is extremely important for determining CPU utilization in conclusion.


"Per clock instructions per clock"? :tongue:

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft
March 29, 2002 5:59:31 PM

I’ll take a stab at this.

Let me start by saying I have nothing against SSE(2) on any streaming multi-media instructions.

SSE is good for doing a bunch of similar things in repetition. It’s very friendly to caches and memory systems and provides great throughput for getting the most done in a short period of time. In terms of complex instructions per clock, it rules the silica. Especially SSE2, which can execute 2 64bit IEEE floating point numbers a tick. The major downside to this is you can’t do anything else while executing in SSE mode. On the other side of things you have the standard out-of-order-multi-staged-pipelined-super-scalar architecture, which can execute a variety of things at once. In terms of all around responsiveness, it rules the silica. Depending on the depth of data, such architecture can hold its own, while getting a head start on operations which may be delayed by the exclusive execution nature of SSE. In terms of programming, the later is essential while the former is a fringe benefit that should not exist without the later.

All errors are undocumented features waiting to be discovered.
Anonymous
a b à CPUs
March 29, 2002 6:05:51 PM

I understand the IPC very well, me I wanted to know as to how the profits of performance SSE2 can be measured per clock.

Forgive my English please, I am of Portugal.
March 29, 2002 6:07:31 PM

Hmm... This is not only a reply to you, Quetzaquatl, but also to Nikko. Actually, it is more to Nikko thqn to you, but never mind.

What I try to say is that in a real life situations IPC is not a constant. Ever heard of pipeline stalls? Cache misses? Different clock cycles per intruction regarding SSE or SSE2 and stuff thelike.
And it is not only the implementation of SSE and SSE2 that is the programmers part of the optimisation process. As I said in other threads, I know somebody who gained up to 40 percents of performance just by changing the order of execution in his SSE assembler code.

Greetz,
Bikeman

<i>Then again, that's just my opinion</i>
Anonymous
a b à CPUs
March 29, 2002 6:09:05 PM

I understand what you be saying now, thank you.

Forgive my English please, I am of Portugal.
March 29, 2002 6:11:32 PM

No prob, and that is why I hate when ppl say X CPU is god in X app, when they don't show proof per clock. Overall performance does not mean Better optimized program. People who see 320 FPS in Quake 3, think that P4 rocks at it and is "optimized". Well per clock, Athlon does more. That is not optimized at all. It just means that if you want to run the best out of Q3 for now, the P4 is better, but not OPTIMIZED.
Thanks again for understanding.

Another example, take the video cards with their new technologies. A Geforce 3 Ti200, vs GF2 Ultra in the new game Aquanox. The game is optimized for new vid cards and gives them so much special effects with little performance drop. The GF2 Ultra is clocked higher than the GF3 Ti200, and also the clock per clock in GPU comparisons does not really vary without the Crossbar memory controller of the GF3, and even yet, the GF3 Ti200 jumps ahead and does not suffer the GF2's consequences in front of major special effects and tons of action on screen. So the people at Massive and Fishtank, did a game optimized indeed becaues at a lower clock the Ti200 with its new technology extensions wins.

--
For the first time, Hookers are hooked on Phonics!!<P ID="edit"><FONT SIZE=-1><EM>Edited by Eden on 03/29/02 03:20 PM.</EM></FONT></P>
March 29, 2002 6:36:54 PM

Nope, I disagree. If an XP2000+ performs equally or sometimes better than a P4 2.2, and Quake 3 runs better, or much better, it means that that app likes the P4 more. I don't know wether ID's engineers did this, but this might have been optimisation. Where does the per clock performance come in here? The only thing you can say, is that you can compare the IPC's of an optimised and non optimesed application. Even then, just comparing the performance on the same sytem would do the thing.
But maybe I'm wrong and I just don't understand what you mean.

Greetz,
Bikeman

<i>Then again, that's just my opinion</i>
March 29, 2002 6:46:26 PM

Actually now I don't get you!
I said that if per clock (take benchs comparing a 1.6GHZ AXP vs a 1.6GHZ NW) the AXP is better at Quake 3, but the overall performance is better for P4, it's cuz it is radically higher in MHZ thus recovering the low IPC. That is why I said people should not confuse Optimized with Overall Performance.

--
For the first time, Hookers are hooked on Phonics!!
March 29, 2002 7:13:00 PM

Don't mean to sound radical, but ... 'per Clock' is dead, every way you look at it. That should be clear by now.

Bikeman

<i>Then again, that's just my opinion</i>
March 29, 2002 7:14:48 PM

i hate people who attack intel for no reason as well, but come on, they do suck dick

But the man in the shop said the via C3 was the best processor money can buy!
!