AMD and Intel Processor differences

[Incroyable HULK]

This is a never ending debate but here is my insight:

I've been active in computer hardware for the last 6 years and I can tell you I still prefer Intel in the long run.

Why??? Chipset stability. Intel has the most stable platform on the market (except for the i820 chipset if some of you remember the SDRAM issue with the PIII).

AMD has done a great job starting with the Athlon lineup a few years ago and I beleive they have the BEST processor out there today. The fact is that the Pentium 4 with his netburst architecture is a total failure. We all knew back then when they launched the P4 at 1.3Ghz and using highly priced Rambus that something was wrong.

Even a PIII 1Ghz had the same performance as the P4 1.5Ghz. And now, the gap is still increasing since a Pentium M (PIII somehow based) running at 2Ghz is able to beat a 3Ghz Pentium 4 (except maybe for Video applications).

Even Intel had to admit :

http://it.slashdot.org/article.pl?sid=05/0...137&tid=218<

AMD saw this comming and they started naming their processor based on a performance comparison instead of the real Mhz Clock.

So here is the deal, go with a AMD CPU on a Intel chipset (i666 platform)

Seriously, if you go with a Intel CPU, go for a Intel chipset otherwise you lose the advantage of a Intel Platform. This is what I buy at work.

If this is a personal computer you are looking for, you should consider a AMD cpu on a nVidia chipset. They have proven to be stable enought. Don't go for ALI & SiS chipset

BUT I must point out that 80% of the PC out there are Intel so guess on which platform your preferred software is designed on (and debugged first).

I compared a AMD64 2Ghz and a Pentium 4 3.4Ghz last summer (2004) and I can tell you they give the SAME performances.

Hope this help

Edited October 31, 2005 by Incroyable HULK

[Thai3g_Eclipse]

Wow, that has to be the first non-biased and not fanboy-ish post I've read. I applaud you sir for making a great post without swaying to either sides. Plus I do also fully agree with you.

[saugatak]

Incroyable HULK, I agree with everything you said but THIS . . .

BUT I must point out that 80% of the PC out there are Intel so guess on which platform your preferred software is designed on (and debugged first).

. . . .with respect to the x86 64 bit chip market ONLY. I agree with the above statement as it applies to the rest of the x86 market other than 64 bit x86 market.

Intel reliability has rested in part on 2 factors:

1. optimized C compiler for Intel; and

2. bug testing done mainly on Intel (factor mentioned by Incroyable Hulk)

BUT the above 2 factors are reversed for the x86 64 bit market, where the C compiler is optimized for AMD 64 bit chips, and M$ 64 bit Windows XP is BUILT for the AMD 64 bit chip.

Now I know Intel reliability has also depended upon quality of chipsets and there can be an argument made for a lot of 3rd party chipsets provided by 3rd party vendors being flaky, but there are also good 3rd party AMD chipsets out there now, particularly for the 64 bit market.

[hbinded]

Maxamoto:

You're an American living in Germany laughing at scared Canadians in Kuwait destroying AMD based machines while figting in Iraq.

Makes sense to send American troops stationed in Germany to Iraq, doesn't it? Ah, it's all coming together for you now? Glad I could clear that up. And yes, the last time I was in Kuwait for OIF the Canadians were there too, and yes, they were all crying like little girls. In fact, no less that 20 countries have soldiers in Kuwait / Iraq, so it sounds like you need to start actually paying attention to the news instead of just watching it

And as far as your little comment about Darwinism goes, this will be my 3rd time in Iraq, and my 6th time in a combat zone. Looks like I win again

Dude! that's a lot of gas coming out there! I'm German! I know much more than I should know about the military cos my dad is in it!!! that's total BS. there is no such thing as american troops stationed in Germany, waiting to pounce on Iraq!!! If I were you, I'd go back to playing RTCW, and pretend it's reality!!

[Wicket20519]

Dude! that's a lot of gas coming out there! I'm German! I know much more than I should know about the military cos my dad is in it!!! that's total BS. there is no such thing as american troops stationed in Germany, waiting to pounce on Iraq!!! If I were you, I'd go back to playing RTCW, and pretend it's reality!!

Er...what?

Anyways - I loved the post at the top of this page. Just wanted to say I thought the author presented a fair view of both processors.

And I also wanted to say I love the new Pentium M's. Heck, even the Celeron M's are nice!

[maxamoto]

Dude! that's a lot of gas coming out there! I'm German! I know much more than I should know about the military cos my dad is in it!!! that's total BS. there is no such thing as american troops stationed in Germany, waiting to pounce on Iraq!!! If I were you, I'd go back to playing RTCW, and pretend it's reality!!

Huh? Have you lost your mind? Here, have a little read, and da Ihr Vater in der Armee ist, sollten Sie dieses bereits gewußt haben!

US Army Europe home page

My Brigade

My Unit History

So, about that RTCW... Maybe you should tear yourself away for just a few minutes, no? Remember: The keyboard is a Tool, not a Toy. Think before you type, and should you fail at that, do us grownups a favor and break both your wrists

Edited October 31, 2005 by maxamoto

[maxamoto]

Uhh, Maxamoto, have you been reading the links posted to? Here's another one:

http://www.theregister.com/2005/10/29/intel_xeon_2009/

FYI, when Intel withdraws its processors that compete with AMD and says it will release new 64 bit chips with a new design and compete with AMD year or two from now, that is an admission on Intel's part that they can't compete.

What did you expect? Intel to actually come out and say, "We suck compared to AMD."

Maxamoto, with respect to your traveling around the world and noticing all the AMD processors failing wherever you go, watching Canadians cry like babies, etc., etc. -- whatever, it's hard to believe.

P.S. I withdraw the personal attack on Max. Went a little too far there, I admit, heh.

Uh, you're sending me a link to an article by the The Register? One of the most self-admitted corporation haters / underdog promoters? Sorry, let's see some scientific facts here, not fanboy sites like Tom's hardware (who accepts a TON of advertising money from AMD, I might add) or Anandtech. Let's see some SCIENCE, kids, not your blind zealot-driven faith.

And what good is 64 bit if it's running 32 bit emulation? Isn't that just as bad as "multithreaded" Windows 98 running all thoses processes down to two threads for DOS to deal with? Like all the AMD zealots are so fond of saying in this forum, faster is not necessarily always better. Folks, I think we can sum this whole discussion up like so:

Q: So, which is better, Intel or AMD?

AMD fan: "DOOD! I get like, a billion freakin' frames per second in Planetside on my new AMD! It ROXXORS!"

Intel fan: "Depends on your application. If you are running mission-critical applications, process-intensive databases and you absolutely require the maximum amount of speed, stability and quality, then Intel would be your logical choice. If you are only interested in gaming and aren't as worried about quality or superior uptime, than AMD will suffice"

Seems to me that, the majority of the pro-Intel camp are working professionals that know how to run more than just a gaming rig. It would also seem that the pro-AMD camp generally comes off like your average junior high hick gang. Sorry, it's always going to be like that. If you like Budweiser, Wal-Mart and spendin' your time at the rodeo in a sweat-stained wifebeater, then definitely, AMD is for you. On the other hand, if you are a working-class professional who appreciates quality, workmanship and can comprehend big words like availability, scalability and price / performance ratio, than Intel can satisfy your need for a higher standard.

Just living up to my reputation as a hyperactive PR agent on damage control. But then again, I live for damage. Cheers!

[saugatak]

Uh, you're sending me a link to an article by the The Register? One of the most self-admitted corporation haters / underdog promoters? Sorry, let's see some scientific facts here, not fanboy sites like Tom's hardware (who accepts a TON of advertising money from AMD, I might add) or Anandtech. Let's see some SCIENCE, kids, not your blind zealot-driven faith.

Step 1: First you demand proof. You get the proof with articles to Tom's Hardware, AnandTech and other sites.

Step 2: After getting the proof, you say, Oh yeah I guess you guys have proof. But that's not real proof. Those are fanboy sites. Now I want "real" proof.

Step 3: Meanwhile, you sit on your a$$ offering as convincing proof for your position . . . YOUR WORD. Wow, that's real convincing evidence.

In any case, if all the testing done by Tom's Hardware and Anandtech (and Anandtech used 64 bit operating systems and databases) and Intel's own admission that they have to scrap their existing "latest" technology because it sucks isn't enough to convince you, nothing will.

Edit: Got rid of non-relevant portions of Maxamoto's quote above.

Edited November 1, 2005 by saugatak

[maxamoto]

DOOD! I get like, a billion freakin' frames per second in Planetside on my new AMD! It ROXXORS!

Nuff' said

[saugatak]

DOOD! I get like, a billion freakin' frames per second in Planetside on my new AMD! It ROXXORS!

Nuff' said

Uh, Dood

Clear the buckshot from your head.

You're quoting yourself there, not me.

See your post #97. Duh.

Edited November 1, 2005 by saugatak

[maxamoto]

Clear the buckshot from your head.

You're quoting yourself there, not me.

Right. Too bad good humor is wasted on bad brains. Let me break it down for you. Here's me a few posts back:

"Seems to me that, the majority of the pro-Intel camp are working professionals that know how to run more than just a gaming rig. It would also seem that the pro-AMD camp generally comes off like your average junior high hick gang."

So, we all know which camp you obviously fall into. So, I was quoting me, making fun of you. Get it now? Man, not only are AMD zealots completely devoid of all common sense, but they lack a sense of humor.

[saugatak]

So, I was quoting me, making fun of you. Get it now? Man, not only are AMD zealots completely devoid of all common sense, but they lack a sense of humor.

GI Schmoe, as I have been complimented on this thread for my biting sense of humor, I beg to differ.

Also, your explanation makes NO SENSE because you put MY NAME by YOUR WORDS. That's called a MISTAKE. Your attempt to explain your dumb mistake as a lame attempt at humor is called PATHETIC.

Finally, I find it weird that I'm called an AMD zealot when all of my machines are either PIII or P-Ms. I run PIII's and P-Ms because I think the P-IV is a piece of crap.

I'm happy with my machines as I use them for common business use and have no need to upgrade. But when drivers for WinXP 64 bit become more common and more programs go native 64-bit, I will upgrade to the best the market has to offer, and that is dual-core AMD x86 compatible 64 bit chips.

By doing so, I'll have saved my business some money in not making any unnecessary hardware upgrades to P-IVs.

If Intel offers a better solution between now and then, I'll go with Intel.

It'd be more appropriate to say, I call it like I see it.

Edited November 1, 2005 by saugatak

[Marsden]

The difference is quite simple. AMD employs Direct Connect which is a direct connection between multicores or multi CPUs and the system memory.

Intel still employs a MCH (Memory Controller Hub) and a slow FSB (Front Side Bus) between cores or multi CPUs and all IO inputs. This is why most Intel CPUs generally clock higher. They have to make up for the slower FSB and MCH constraints.

[Martin L]

The next one who doesn't like one or another and expresses this on this forum will have a problem....

[suryad]

I thought the reason that Intel had problems with the P4 not scaling well with the clock speed increases besides the memory controller was the fact that it had some insane number of stages in the pipeline.....31 to be exact in the Prescott....and apparently longer pipelines enables people to crank the clockspeeds....so since Intel was marketing morons with the idea that higher clockspeed equals more speed...they just kept increasing the number of stages.