P4 1.8GHz or AMD Athlon XP And Why?


Recommended Posts

One thing that everyone seems to forget, especially the AMD supporters is that Intel has abandoned the x87 instruction set in lieu of "Netburst.". The AMD is still mired deeply in that x87 set. Trying to benchmark these two CPU's on an even playing field is impossible because of this. Most of the current benchmarks are weighted heavily in favor of x87 and as a result the Intel processors find themseleves at a distinct disadvantage. A 1.5GHz P4 will run x87 at about the same proficiency as a PIII 850.

It's like trying to benchmark an eagle against a barracuda, each will outperform the other in it's defined environment. Unfortunately, there is not enough software written that takes advantage of Intel's new instruction set, so they do in fact run slower on the current software/benchmarks.

I would still take an Intel processor over an AMD simply because Intel has been in the chipset business for quite some time. The 440BX was one of the longest running, fastest and most stable sets ever made. Intel knows how to optimize it's processors with companion chipsets.

Further, heat is an issue. I have been building and servicing PC's for 7 years and I can tell you two things based on my experiences.

1. AMD systems are not as stable.

2. Very few people ask for AMD by name compared to those who only want Intel.

I read here that reviews on sites such as Tom's Hardware prove this or that. Well Tom's Hardware is the least bias site on the Internet when it comes to this issue. For example, not too long ago they did a Celeron/T vs. Duron comparison.

http://www6.tomshardware.com/cpu/02q1/0201...0103/index.html

If you read the set up closely, the mainboard they placed the Celeron on did not have the ability to utilize PC133 memory, yet the Duron's did. The ensuing results were slightly in favor of the Duron, but with a better board I have no doubt the Celeron would have prevailed or at the very least been more competitive.

Last week, I built a Celeron 1.2 Tualatin, on an Abit ST6 board with 256MB of PC133 and a 64MB GeForce 2 GTS Pro Video card. It ran just under 7000 in 3DMark 2000. I thought that was pretty damn fast for a $99 processor on a $100 mainboard with an older Video Card.

In any event, this arguement will never be settled.

I for one choose Intel. And a newbie is the last thing I am.

Krome posted:

       

Most AMD user wants power. Most AMD users are very young and new to computers. In reality, if you want power, you get power but you will loose quality. AMD users does not understand what quality is. To them, power is speed, speed is power. That's all they care. Don't talk quality to them. They'll make you think that power is their quality of their CPU. Yes. But quality and reliability has to be equal out or you else you don't have quality. They are young. It will take times for them to understand everything. The law of the universe is if you want the best of one thing, you have to sacrifice another. That's what AMD chip is all about. They neglect reliability over quality. I am not surprise if AMD user wouldn't understand my wide-spectrum of logical reasoning. Once you are fixed to something, it's hard for you to give it up. What I would be surprise is if there's Intel (Pentium) users that are very young. This is because they are smarter than average.

I have some questions for you;

1.) Why do you think that AMD does not produce quality chipsets? Ever since then K7 series hit the retail shelves AMD chips have been extremly reliable. Yes, there were compatibility problems with some bios, some gpus' and certain installs of Windows. But that was well over a year ago. At the same time AMD was having these compatibility issues, Intel was also going through a wide series of motherboard issues. This became so bad that a large number of retailers dropped their Intel certification (roughly 15%) because compatability issues with Intel chips were costing them too much money. I am responsible for a moderate (20+) number of machines, and as each Intel chip goes down, an AMD takes its place. Over 2 years that's been 10 chips. (Which is nice because that means I stay pretty current on the latest chipset...)

2.) Power doesn't equal quality, but don't you think it has something to do with it? I take from your post that you would disregard a chipsets overall performance in determining quality. This absolutley must be included in determining overall chip quality. The time and cost of fixing any incompatability issues is in direct relation to the money an employeer loses by deploying slow workstations. Render time, save time, network time (computing time) equals hard cash when that time is veiwed over the long term. Dollars a day waiting on a slow processor becomes thousands of dollars when looking at tens/ hundreds of employees over weeks/months. With AMD & Intel having almost zero difference in compatibility, AMD does process faster than Intel, hands down. In cost of deployment and long term production, AMD is the clear winner in the x86 field. (I love/ hate Macs and they run a slightly better dollar value than AMD based PC's in terms of long term production, but that's another discussion.)

3.) What do you think about the current roadmap of Intel vs. AMD? It was announced several months ago that both Intel and AMD are working on new bus systems to replace our outdated standard. The difference being that AMD is establishing these standards in line with a consortium of Mobo manuf., while Intel is going at it alone. AMD's system is supposed to support legacy bus cards, such as PCI, while Intel's won't. Hmmm. See any problems for Intel coming?

Even the quality of the chipset design roadmap comes into question here. Why did Intel revamp the P4 to handle instructions like it does? For the clock speeds alone? Out of all the chipmakers, Intel is still riding high on the misconception (read: taking advantage of public) that clockspeed=power. AMD is expanding the chipsets pipeline bandwidth, is running wild with DDR, and is designing their chips to handle instruction sets more like Motorola does (kinda - this is hard to describe non-technically -> help?), which will be far more efficient. (Although if Intel goes through with hyper-threading they will be so very cool I'll scream, AMD...watch your back)

4.) Please explain exactly what your "wide-spectrum of logical reasoning" is. Even my 4yr. old nephew, who seen Star Trek, knows logic is based upon factual reasoning. Abstract logic, systems logic...all logic references from some factual point. I missed yours. I found a few opinions, none of which were backed by statistics, experience or any technical information.

I am an immature AMD supporter. They won me over with their cool green logo and really insane processing power. Wow. I AM FRAGBAIT ! (with 10 yrs. experience in IT...)

Originally posted by obsoleet  

for not being a newbie you should know that there is no such thing as a celeron tulatin...it's the Pentium III-M or P3 tulatin.

What the heck are you talking about man? All of the new Celerons are Tulatins. They require a FC-PGA2 compatible chipset just like a PIII Tulatin.

DeadZombie:

Legacy? Do you still have an ISA slot in your AMD Athlon 2000? I didn't think so. Legacy is a BOTTLENECK.

Soma:

"Opinions mean nothing to me... It was once an opinion that the sun revolved around the earth, it was once an opinion that the earth was flat, and it was once an opinion that by allowing someone to nearly bleed to death, you could drain them of a certain disease or sickness they had... "

Actually, none of these were based on opinions, but what were considered "fact" at the time. In fact, the opinion that the sun revolved around the earth (Ptolemy's Earth-centered view) was mathematically based and indeed DID work (until new findings arose). This has nothing to do with this conversation, but I thought you might find it interesting.

I'm sorry, if i knew you judged this type of thing on how well it runs without the fan (BTW, the Athlon XP runs as well and shuts down unlike the Athlons before it) i'd say GO PENTIUM!

Sorry. But when a slower AMD Athlon XP beats a much faster Pentium 4 in 9 out of 10 benchmarks, for half the price, I go for the best one -- the Athlon XP.

toms did benchmarks. hard did benchmarks. I did benchmarks. anand did some benchmarks. tweak did some benchmarks.

Take 10 benchmarks comparing both you'll see the XP is equal or BETTER than the Pentium 4.

Just my $.052 (Inflation rates)

I don't know why you guys keep bringing this up.

Both chips use a different clocking scheme.

Simple as that, you can't compare point for point.

If you really compared point for point than consider that an 800 mhz MACINTOSH can kick the living crap out of any AMD invented thus far.

I read about the first 2 1/2 pages of this so if this seems out of place you know

Just my 2 cents, but here are the advantages to the Athlon:

9 stage pipeline (PowerPC G4s have a 7-10 stage, smaller pipeline means more IPC but lower clock speed)

NVIDIA nForce chipset!!

Cheaper

MP capable out of the box

Easily OCed

Intel Disadvantages:

20 stage pipeline (very low IPC so much so that the clock frequency must be around 3.6ghz to make it worth it)

Poor DDR support

RDRAM is highspeed but not very wide being only 16 bit

QDR FSB, thats only 100mhz core, AMD has a 133 to get DDR 266

Go with Athlon because it perfectly balances IPC and MHZ to give you maximum performance at a better price.

And for you Intel guys, its sad that a 1.67ghz part can beat a 2.0ghz and meet a 2.2ghz

Heinrich-X; Right On! Don't forget if the pipeline needs to be flushed the new P4's take about 4 times longer than an Athlon XP.

and oh yeah, AMD chips don't overheat in properly built systems.

That's like saying your HD doesn't work because you use it to hold your refrigerator magnets.

you guys are ****ing stupid. I have an athlon system i built not too long ago and it has endless heat probs and **** . I use a thermaltake Volcano 5 HSF and arctic silver II and the damn thing still hits 50 C! My P4 on the other hand runs at 28C all the time pretty much and whoops my athlon system to hell :evil:

volcano 5 is a **** heat sink and are you sure u applied the artic silver correctly??

too much artic silver and it will raise the temp rather then cool it

i dont have a prob with my XP

u probbably dont know how to build computers or something i dont see any other thing it could be

listen when u put on the artic silver 3 make sure its a paper thin layer not tooo much!!!

also thing about getting yourself a better heatsink

and also some system fans not for your cpu but for peripherals my mates Intel CPU used to be ok temp but his case insides was very very hot

get yourself some fans i went over the top but this is my case for my AMD XP 2000 my mate with the intels case is more or less the same

we wanted it coz it looks cool and also for some serious overclock power

the window mod lights up to neon blue in the dark

Rockin69-2.jpeg

Rockin69-3.jpeg

the inside is very cool had to do a simillar set up to mates Intel

if u dont like noise then get yourself a speed adjuster contoller it switches fans off 1 by one but if u want sheer overclocking power as my mate and i then u need to keep system cool with fans

not as many as mines 1 or 2 will do the job if your not gonna overclock

You know what's funny also.. some of these kids have a so-called "kickin'-Intel's-ass" processor, but then they own something like a 56k connection.. talk about using that PC to get more FPS! ..like it would matter.. but good points put out by Freeza, hats off to him :)

and by the way irdawood, could you heat me up some hotpockets?

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • I've owned nothing but ATi/AMD GPUs since 2002, after my last nVidia GPU in 2001 (3dfx before that), IIRC, and in all of that time I recall getting this error maybe once, certainly no more than twice. Despite all the scuttlebutt as to how poor AMD drivers are supposed to be that has certainly not been my experience at all... Usually it has been a configuration problem of some kind. Then again, since we're dealing with OS versions that are EOL, it could easily be an OS version discrepancy. It's still weird to think that Win11 has been officially out for more than five years!
    • AI will never be the jobs panacea some companies fantasize about today. Oracle is likely using it as an excuse, which we will see a lot of companies doing, I'm certain. They love their "plausible" excuses for their downturns. A couple of weeks ago my wife asked me to call Krogers about some discrepancy in a online grocery order, and it will be the last time either of us does that. I'll just do emails with humans from now on... The AI experience was horrible--the obviously recorded voice started asking a bunch of questions about our orders six months prior(!) and saying, "Is this in reference to your order on January 6, for $****?" You say "No!" and immediately the next question is "Is this in reference to your order on January 29th, for $****?" again, I answered "No!"--and it was incredible--on and on it went like that for fully 20 minutes until we finally got to the present, and only then was I put through to a human with authentic intelligence... I wondered why on Earth the idiot AI didn't start with the most recent orders and work back from there, as it was something anyone with a functioning brain would have done. And why didn't the AI have enough sense to ask me what the problem was in the first place? It didn't take too much deduction to understand that the goal of this "AI" was to cause the person on the phone to hang up in disgust, with no resolution of the problem. That begs another question: why pay for a tool-free problem line if the goal is to avoid solving your customer's problems?... Fortunately, Krogers does have real humans capable of reading an email and understanding it, and if she sees another situation in the future that's route she or I will take. The online grocery delivery service from Krogers has been great, over all, but their AI truly sucks.
    • AI is the justification that company administrators use to lay people off; it is not the end all, be all touted in the media (many of whom can't tell a microchip from a potato chip). Greed is main driving factor behind its adoption; the other is remaining relevant in the face of competition from other entities.
    • Firefox 152.0.2 is out with fixes for performance, translation, and cloud storage services by Taras Buria A new bug-fixing update is now rolling out to Firefox users in the Release Channel. Less than a week ago, Mozilla fixed crashes on Intel Raptor Lake processors with version 152.0.1. Now, Mozilla has prepared yet another set of fixes that address problems with localization, playback issues of certain MP4 files, and performance issues on website that perform various encryption operations at once. Here is the full changelog: Firefox 152.0.2 is now available for download from Mozilla's FTP. Existing installations will get the update over the next several hours. The latest version will also be available soon on the official website, the Microsoft Store, and Neowin's Software page. You can find Firefox 152.0.2 release notes in the official documentation. In case you missed it, Mozilla released Firefox 152 earlier this month. The latest feature update brought reworked settings with a more streamlined user interface, JPEG XL support, new features for Private mode, a new way to mute a tab (just type "mute" in the address bar), and many more. You can find the complete changelog here. In other Firefox news, Mozilla recently published its roadmap, where the company detailed the upcoming Nova redesign and other features it plans to implement. Mozilla wants to make the new user interface easier to navigate and more modern, with a heavy focus on its privacy tools, such as its built-in VPN. If you are curious, you can already enable the new UI as described here.
  • Recent Achievements

    • One Month Later
      timbobit earned a badge
      One Month Later
    • One Month Later
      nates earned a badge
      One Month Later
    • Week One Done
      Almohandis earned a badge
      Week One Done
    • Rookie
      dorf went up a rank
      Rookie
    • First Post
      mike_rumble earned a badge
      First Post
  • Popular Contributors

    1. 1
      +primortal
      481
    2. 2
      +Edouard
      172
    3. 3
      PsYcHoKiLLa
      103
    4. 4
      Michael Scrip
      88
    5. 5
      neufuse
      70
  • Tell a friend

    Love Neowin? Tell a friend!