The Great Debate: AMD vs. Intel

[Syborg1211]

Hmm, well, Intel has at least one more factory about to be completed, not sure if there are more, but i am guessing so. What does another factory mean? More availability, and in turn, lower price because of more availability. P4 should be quite an interesting chip, we shall just have to see how the P4 and Mustang duke it out. It is hard to say who will be the best since most people's "facts" are fictional.

cant or dont feel like answering #1 and #2

3) I prefer the P3 mainly because i already have a motherboard for my chip and this lessons the costs of a new system. And... this is a big one. There really is NO price difference!!!!!!!! Once you think about it, the price of a motherboard for an amd chip is much more than one for an intel chip, and therefore makes the lower costing amd chip not so lower costing for people building new systems. But once you get the motherboard, you can keep that motherboard and buy a new cpu.

 

[rockhard]

Heh, quad pumped system bus and DDR ram.
Sounds like a overclockers nightmare a brewing

 

[Modus]

Red Dawn,

<<Yeah you qualify as a judge for that Zealot.>>

There can be no &quot;judge&quot; of an objective fact. I was merely pointing out that all the replies to the thread up till that point had been grounded in factual evidence, and that certain posters had presented the facts in a particularly clear and coherent manner.

The fact is that AMD systems currently offer better performance for less money than Intel systems. Even taking into account the cost differences of motherboards, this cannot be disputed.

Ironically, your absurd portrait of AMD chips as &quot;overheating rockets&quot; puts you solidly in the camp of the zealot, into which you so flipantly and enthusiastically toss me. In fact, there has not been a single occurance of overheating (ie. measurably affecting system stability) that was not solved by proper adherance to the simple, easily accessible and cheaply applied cooling recommendations set forth by AMD, or by the RMA of a clearly deffective and anomalous part. Your mention of &quot;AMD's heating troubles&quot; in the same breath as the old pro-Intel dogma of &quot;poor motherboard support&quot; simply amounts to the same recycled Fear Uncertainty and Doubt (FOD) rhetoric that current Intel disciples are forced to fall back on in the face of scant rational support for their misguided loyalties.

Basically, the only interesting thing about your post was the childish namecalling. And that's sad.

Modus

 

[rockhard]

Modus

hehe, hope this aint goin to be the start of something

Its funny how people have preconceived ideas and dont make decisions at face value.

What really tipped it for me was the intel systems ive had in the past (BX6r2, BE6) left a sour taste in my mouth due to just goin with the crowd.
I did the same with the P3V4X but struck lucky that time
Thought it was time to give the opposition a chance so to speak.
What i really like about the T-Birds at the moment is how they arent as sensitive to heat as the P3's
Theres an article over at [H]ard OCP where they gunned a P3 1ghz and a T-Bird 1ghz.
The T-bird did 1200mhz with 1.8v ) and they reckoned that it was voltage holding it back not heat? hmmm? Was as stable with air cooling as with H2O cooling. Now i know they are biased gettin theyre samples offered to them by the manufacturers, but seein as both Intel and AMD had it in their interests to offer their best parts was a pretty even playing field in my book.
Now the thought of overclocking a 1ghz P3 at over £800 or a T-Bird at just over £400 is a no brainer for me
Also when i read the newsgroups fraid all the current motherboards are having one problem or another. That leaves unbiased evaluation of the scene the only deciding factor in my book.
The thought of buying another P3 and trying to gun it this time round felt less appealing. Just take the plunge and buy a gig in the first place, then u wont be dissapointed. Any extra comes as a bonus now
Now i wonder what my 1ghz T-Bird goin to do when the alpha arrives? Heh, the joys of overclocking

Ooops got all carried away again
Sorry :/

rockhard =)

 

[rockhard]

This threads digressed a bit from what the_querist wanted to know, so heres some info on the P3 and T-Bird ripped from another site:-

Pentium III 1GHz Specifications

· 29 million transistor 0.18-micron Coppermine core
· 1GHz clock speed ? 7.5x clock multiplier
· 32KB on-die L1 cache running at core speed
· 256-bit Advanced Transfer Cache - 256KB on-die L2 cache running at core speed
· Advanced System Buffering
· 242-pin Slot-1 GTL+ CPU interface running at 133MHz
· 1.70v core voltage


The 1GHz Athlon &quot;Thunderbird&quot;

· 37 million transistor 0.18-micron Thunderbird core with Copper interconnects
· 1GHz (1000MHz) clock speed - 10.0x clock multiplier
· 128KB on-die L1 cache running at core speed
· 64-bit exclusive 256KB on-die L2 cache running at clock speed
· 462-pin Socket A EV6 CPU interface running at 100MHz DDR (effectively 200MHz)
· 1.75v core voltage

Maybe discussin the implications of the differences between the 2 be more apt

Chew on that guys,

rockhard =)

 

[Modus]

Rockhard,

There's certainly nothing wrong with going against the crowd, supporting the underdog, etc. I actually support that kind of sentiment. But these days, one need not be loyal to AMD or naturally supportive of the underdog to realize that AMD systems are simply a better value. It's been shown time and again in benchmarks by respected authorities that AMD processors are, on balance, at least as fast as equivalent Intel processors, and often significantly faster (Duron vs Celeron). Couple that with their clear price advantage (even granting slightly more expensive motherboards) and their stability and ease of use (relatively, considering the break-neak pace of modern PC hardware), and you've got a clear winner. It doesn't take zealotry or loyalism to see that.

Modus

 

[chiwawa626]

Screw AMD VS Intel, Howbout RISC VS. CISC, or better yet, PC Vs. MAC

 

[rockhard]

Modus

Couldnt agree with u more

When i get my rig runnin at its max lookin forward to adding my benches to my sig just to see how many intel zealots can come up with benches that can compete

The way my benches at the moment are lookin goin to be real interesting to see what a NV20 goin to look like on my rig

Wish nVidia would hurry up and get it rolled out. My rigs just screeming for it

Its 3am here in the UK and i need some sleep, so gotta go fraid

Keep yanking at them zealots legs for us

Cyas around,

rockhard =)

 

[Modus]

Red Dawn,

<<Modus, you've were an AMD Zealot before it was fashionable to be one.>>

I have no axe to grind. I actually advocated the Celeron 300A in its day (who wouldn't?). A zealot is some one who backs up their arguments with emotion instead of reason and fact. Either show where I made false claims without factual support, or quit this petty name calling and stick to the issues.

<<The truth is the Motherboards right now aren't as stable as we'd all like them to be.>>

As an Elite Member, you shouldn't be spreading this kind of FUD. Socket A (KT133) is an extremely minor derivative of the solid, stable Slot A (KX133/AMD750) platform that's been out for months. Socket A is not terribly new, nor is it &quot;broken.&quot; I've not seen a single Socket A stability problem on this board that was not resolved by the simple procedures relevant to ALL platforms: installing correct drivers, updating the BIOS, using aproved cooling methods, or RMA'ing a defective board. And remember, there is no greater testament to the exceptional stability of the Socket A platform than the fact that Anand himself is using them to power these very forums.

<<The only difference between you and GUTB is that GUTB was an Intel Zealot.>>

Blah blah blah. . . name calling. . . yadda yadda yadda. . . dredging up GUTB. . . blah blah blah. . . personal attacks. *sigh* Argue the issues, or go home.

<<Price performance we've already debated.>>

And we came to the conculsion -- which you so quickly gloss over -- that AMD systems are always cheaper than equivalent Intel systems, by a margin that ranges from small to huge. Let's not forget that, in the absence of any proven, methodic study of stability and reliability, we must base our entire decision on price/performance ratios. Translation: AMD, for now.

<<I say if you want to be safe and sure that you will get a stable Motherboard CPU, either go the way of a Slot A Tbird or a P3 and a BX or Via Chipset.>>

I agree. If you value stability above all else, go with a platform at least one generation behind.

<<I'm sorry, I know it isn't popular to speak out againts the Socket A Platform, but damn it, what I am saying isn't BS.>>

Well what do you call claims that AMD chips are constantly overheating, that Socket A is inherently unstable, that the motherboards are problematic? None of this has been expressed by any respected industry source such as AnandTech or Tom's Hardware. It just has no basis -- more FUD.

Modus

 

[jmcoreymv]

It used to be there were too many Intel zealots, now the tables have turned, and there are far too many AMD zealots.

 

[Modus]

Why must people brand those on the other side of an argument as somehow different from themselves?

If one person says that AMD processors offer a better value and the other says that Intel processors offer a better value, then one person is right and the other is wrong. Whether those statements are motivated by mistaken assumptions, misinformation, or &quot;zealotry&quot; is entirely irrelevant to the issue. Namecalling is just a mental cruch.

Modus

 

[Modus]

Red Dawn,

<<Yeah right Modus, you mean to tell me that all those having problems with their Socket A Motherboards just need better cooling, a Bios Flash, drivers installed properly , or have bad Motherboards>>

BINGO!

<<There isn't one IT Proffesional that bases his purchases on what either Tom or Anand.. for that matter, what any Hardware enthusiasts website recommends.>>

If so, then the majority of IT professionals are supremely ignorant, because price/performance is right now the ONLY valid buying criteria, and it is reported to near perfection by the hardware community. Brand loyalty and anecdotal reliability reports are not, and can never be, valid methods of selecting computer hardware.

<<I know what I see and what I see is a bunch of people pulling their hair out trying to get &quot;some&quot; of these boards to work stabily or to work at all.>>

*Yawn* more FUD. This really gets tiresome. Show me a single Socket A stability problem not resolved in a timely manner by the usual means of motherboard troubleshooting: BIOS updates, driver installation, cooling recomendations, RMA.

Besides, did you ever stop to think that the reason you see so many AMD troubleshooting posts these days (other than your own wishful thinking) is simply that AMD systems are currently much more popular among the educated kind of people who visit this forum? And since more people are building AMD systems than Intel systems, more people will report problems with AMD systems than with Intel systems. This is a well known effect in any consumer market, and it vividly illustrates why we simply cannot rely on our own perception of anecdotal evidence. The only valid evidence is a statistically accurate survey. I suggest you do one right now -- &quot;Are Socket A users are satisfied with their systems?&quot; The results would prove interesting

<<And regarding the heat issue, hell fire up a Duron or an Tbird for a minute without the fan on the Heat Sink plugged in and you'll end up with a dead burned out CPU. That would be the users fault and I really don't blame AMD.>>

Um, yeah. . . and fire up the power supply set to 230V instead of 115V and you'll end up with the same thing. . . so what? If it's the user's fault, then it's entirely irrelevant to this discussion, begging the question of why you would mention something like that in the first place, other than to spread more FUD.

<<And you better know what you are doing or you will most definately end up with something toasted, smashed or find yourself posting in forums like this looking for answers..>>

More FUD.

Modus

 

[Modus]

Red Dawn,

<<Modus, you are about as stubborn as they get.>>

It takes two to be stubborn

<<Why in the hell should we do doing all this motherboard trouble shooting?>>

&quot;All this&quot; motherboard troubleshooting amounts to three things:

- Install the one (1) VIA service driver that your manual told you to install.

- Ensure you have the most recent BIOS flash update, for bonus performance and stability. Use the default, conservative settings unless you're sure you know what you're doing.

- Ensure that you follow the cooling guidelines set forth by AMD. (Use a cheap approved HSF, apply a thin coat of thermal grease, and properly seat the unit).

Nice try, but all the exageration in the world can't make us agree with your assertion that these simple measures, common to ALL platforms, are too much of a burden. You still haven't taken up my offer to conduct a systematic survey: &quot;Are you satisfied with your Socket A system?&quot;

<<if one has to jump through rings of fire to get their computer running>>

Have you even BUILT a Socket A system? It is SO easy! It's childishly easy. It is NO different from any other platform. The things that make a Socket A systems run smoothly are the SAME things that make any other system run smoothly: drivers, BIOS, cooling.

Your argument is worthless and desperate. Socket A systems are fast, cheap, and solid. End of discussion.

Modus

 

[Modus]

Red Dawn,

<<My isn't that reminiscent of GUTB.>>

*Sigh* Insult what you can't accept, eh?

<<Look all I am really saying right now is that the socket A boards are a little more troublesome than they should be.>>

And all I am saying is that you are wrong. There is absolutely nothing concrete to say that Socket A systems are inherently troublesome, or that they require any special attention not given to other platforms. The number of Socket A related posts on this board simply confirm its massive popularity.

Red, you've been backtracking this whole thread. First it was the insane idea that Socket A chips are &quot;overheating rockets,&quot; then you abandonned that and tried to say the motherboards weren't stable, and now you just claim they are &quot;more troublesome to setup&quot;. Make up your mind which rubbish to promote.

Modus

 

[Osangar]

<< <<There isn't one IT Proffesional that bases his purchases on what either Tom or Anand.. for that matter, what any Hardware enthusiasts website recommends.>>

If so, then the majority of IT professionals are supremely ignorant, because price/performance is right now the ONLY valid buying criteria, and it is reported to near perfection by the hardware community. Brand loyalty and anecdotal reliability reports are not, and can never be, valid methods of selecting computer hardware.
>>



Price/performance doesn?t mean much in the corporate IT world. Of course you always want to get a good price, but? Support costs are much higher then hardware costs. Cost of lost productivity is much higher then support costs. (Software also costs more then hardware)

Even differences in reliability too small for a home user to notice can add up to big bucks in a large organization. Intel has historically had a much lower total cost of ownership then AMD and there is no evidence that this has changed.

This doesn?t really have much to do with home users though. As a result AMD does much better in the home then in business markets. AMD has made some efforts recently to make their product more suitable for corporate IT but I think it will take several years to pay off.

 

[Osangar]

<< 1) Is one really &quot;better&quot; than the other or is it more application specific? >>



AMD is very competitive with Intel right now. In most cases you would want to consider your specific situation and planned use of the system. AMD would probably be my choice for a home system right now, but this depends on lots of things and may change at any time.



<< 2) What are the major achitechture differences between the two (details)? >>



There are a lot of differences but I think the main ones are

The Athlon has 3 integer units while the PIII only has 2. If the Athlon can find enough instructions that can be done in parallel it can be up to 50 % faster then the PIII. Since it doesn?t get this in practice it may be that it doesn?t work as hard at finding this parallelism.
One possible reason is that to keep clock speed high with a shorter pipeline they dedicated less logic to finding parallelism.

The Athlon has a much longer out of order window, which should help it find more parallelism. It also supports more outstanding reads and writes, which should also help.

Both the Athlon and PIII have 1 ADD and 1 Mult floating point pipeline. The PIII shares a port for issuing instructions to the floating point pipelines, while the Athlon shares them with the ALU. The end result is that the Athlon can often execute 2 floating pint instructions per cycle while the PIII is limited to 1.

The Athlon has 64 KB L1 Data and Instruction caches while the PIII only has 16 KB

The Athlon has a 100 MHz DDR front side bus (= 200 MHz) while the PIII has a 133 MHz. (This may seam like an advantage for the Athlon, but it can force the chipset to have higher latency and sometimes reduce performance.)

Despite the advantages the Athlon has the PIII consistently outperforms it on SPEC (the de-facto industry standard benchmark for CPU performance) This is probably due in large part to better software and system support.




<< 3) What do you prefer and WHY? Be specific. >>



There are jus too many variables for me to answer this, but as a company I prefer Intel since I view them as the force driving innovation in the X86 world. AMD makes good processors right now but they tend to play it safe and stay with proven technologies. If Intel took this approach X86 would have fallen to the RISC systems years ago.

 

[noxipoo]

Buy a AMD Tbird and/a Duron and take your chances. Buy an Intel and you are guaranteed that it will work. Plain and simple. End of discussion

BS. then why would anyone buy an AMD chip? and what about the Intel recalls?

 

[Mem]

I feel I must reply to Red Dawn statements &quot;with all due respect&quot;, first my KT133 board(MSI K7T PRO) is 3 weeks old before this I had a Intel based system anyway going over to AMD system was very smooth ,infact I did not even have to format my harddrive ,game wise everything is rock stable(using a Geforce2 MX)&amp; the only thing I did beforehand was do my homework on AMD approved parts like PSU,heatsinks etc this was not hard in itself.

I feel the best judge of AMD &amp; KT133 boards are those that own them while it`s true some people have had problems I still get friends with Intel based systems asking for help &amp; to blame the hardware straight away is unfair.To be honest I cannot believe how STABLE my system is, I to have read post like this before I went over to AMD &amp; thought the only way to find out the truth is to either buy or use one of those AMD &amp; Via boards &amp; see what happens, well I put my money where my mouth is &amp; am very happy,btw also I`m a gamer so stability is at the top of the list.

 

[rockhard]

hehe, still rumblin

How about a Sandra Memory bench of 560ish/655ish?
Thats what im gettin with 147mhz ram setting @ 2,2,2,Turbo
When my mushkin rev2 comes in a day or two only gonna get higher as my EV6 bus seems quite happy to do 120fsb (240) but my crucial memory doesnt

Now can anyone show me a Intel based system doing those kind of benches in Sandra? I know its synthetic but hell, makes a big difference in games compared to my Intel rig

rockhard =)

 

[formulav8]

Red Dawn. Do you own a socket a platform. If not you have no room to be talkin. I have one and it is plenty stable even overclocked. If you are going by the posts on this board then you are about stupid. Most of the people that make claims of amd sucks is because they are to stupid to put a comp together. If they can't get the chip to oc to a 1200 then they say amd sucks. It is because alot of people without exp. trying to put something together.

 

[rockhard]

fomulav8 well put I just didnt have the guts to say it

I think people will always find that when a type of system or specific mobo/CPU is more popular than the others it is going to lead to more post complaints than any other - simple maths

My lil P3V4X was such a case - when it was selling shed loads there was tons of complaints. How many do u see now? Due to not being as popular now and the maturing of the bios etc as mentioned in one of my earlier posts

Why oh why cant these Intel biased people look at it from their own self interested point of view (cash$/performance). Is Intel/AMD going to send them a Christmas card and thank them? Nope, just the suited fat cats are goin to get fatter on theyre lemming style attitude.
To go with the best at the time of ur upgrade is the way to go, and will help to keep these fat cats on their toes and keep the prices competitive

Must go, got a lemming post reply in my mail box

rockhard =)

 

[the_querist]

I just wanted to thank you guys for the good info. Many of the posts have answered some questions I've had about the differences b/w AMD and Intel processors.

PS - I like the idea about RISC vs. CISC and PC vs. Mac - anyone care to start a new thread and take a stab at it? =)

Regards,

the_querist

 

[Mem]

Red Dawn ,If you want the best stability definetly buy a MSI K7T PRO board,Btw if it`s good enough for Anandtech &amp; I,it should be good enough for you,You may miss all the overclocking extra`s of the Abit but you cannot always have everything in life,(well not just yet).

 

[Modus]

Osangar,

<<Intel has historically had a much lower total cost of ownership then AMD and there is no evidence that this has changed.>>

False. There is no evidence that either company has a &quot;much lower&quot; total cost of ownership, period. You are working under the mistaken assumption -- as many in the corporate IT world do -- that Intel processors are inherently and historically better, and that AMD must somehow prove to you that it can match Intel's lofty level of stability and reliability. In fact, there have been absolutely no controlled studies of the reliability of either platform. So, in the absence of reproducable results on stability and reliability, we must eliminate them entirely as buying criteria, or else resort to the kind of hackneyed ravings by people like Red Dawn, Alnoor, and GUTB, who take their own limitted personal experience and expand it to encompass the uniserve, declaring that whatever happened to them must be the norm.

Red Dawn,

<<Still I have also received emails from other who have had a few problems but were afraid to post about it out of fear of suffering from Moduses Wraith.>>

You make me sick. Now you're just resorting to outright lies. Oh and by the way, a wraith is a vinidictive ghost, not wrath, which is fury. Slightly different words. Comprende?

<<Buy a AMD Tbird and/a Duron and take your chances. Buy an Intel and you are guaranteed that it will work. Plain and simple. End of discussion.>>

Buy an Intel and you are guaranteed that it will work? Guaranteed? Do I even have to say anything to that? That's the nice thing about arguing with you, Red. It's often enough just to quote a sentence and bring to attention the sheer, mind numbing stupidity of it.

Mem, Rockhard, Formulav8,

A valiant effort, but most likely in vain. Red's mind is made up. Don't try to confuse him with the facts

Modus

------
&quot;Buy and Intel and you are guaranteed it will work.&quot; -- Red Dawn

 

[Thorn]

<< Price/performance doesn?t mean much in the corporate IT world. Of course you always want to get a good price, but? Support costs are much higher then hardware costs. Cost of lost productivity is much higher then support costs. (Software also costs more then hardware)

Even differences in reliability too small for a home user to notice can add up to big bucks in a large organization. Intel has historically had a much lower total cost of ownership then AMD and there is no evidence that this has changed.

This doesn?t really have much to do with home users though. As a result AMD does much better in the home then in business markets. AMD has made some efforts recently to make their product more suitable for corporate IT but I think it will take several years to pay off. >>



Very true, and well put. Most of my customers could care less if they save a couple hundred bucks on a system as long as productivity losses are kept to a minimum. IT guys get really POed when they get their asses chewed because of down-time.

We are passionate about PC because we love them, it's our way of life. But, most buyers and corporate types think of PCs as appliances (like a fax machine, coffee maker, or copier). The main difference is that PCs hold the most valuable property many modern businesses own, their data/information. A $2000 PC doesn't mean a hill of poop compared to the $500,000 worth of POs, customer info, and product info kept inside it. AMD just doesn't have the track record Intel enjoys and this isn't going to change for some time yet. I believe it will change, they're just not there yet.

I love AMD, and I think they make the best CPUs in the industry. I even like the new Socket A mobos and have sold more than my share of A7Vs. But I don't generally recommend the Socket A platform to corporate customers, with the exception of CAD/CAM operators. I'm not going to risk losing a customer because of a BIOS or driver issue, no matter how remote the possibility (it's also the main reason I exclusively sell Asus to corporate customers). Most times they don't care if they save $200 /PC, they just want the most stable product they can buy and right now that's Intel (CuMine w/ 815 or BX). Now for home users, I always sell AMD unless Intel is specified... Price/Performance makes all the difference to them and I want them to get the most bang for their buck.

It's simple economics on both sides of the fence, and we benefit from the competition. Hell, even RAM prices are dropping again.