What CPU Do You Use?

serAph

uhhhh

this is hilarious. If an AMD xp 2400 (2.0ghz) slaps a 2.4Ghz P4 silly in speed comparisons, how is it that the Athlon is the worse chip? LMAO

tell ya what - I think you're the newbie

mrdarkside2k3

I jus wanna say that I am currently on an AMD Duron 900MHz, and I am gonna upgrade it to at least a 1700+ soon. I will stick with my AMD

__________________
~Rosebud~

Chaintech 7NJL4, AMD Athlon 1800+ @2400+, Radeon 9000 Pro 64meg, (Dual Monitors and TV running), 512 Meg DDR Ram (all Generic), 6 Gig Maxtor HD, 80 Gig WD HD, SB Live! 5.1, BTC 40x12x48, WinXP Pro

KNIBBO

I see advantages in both, but it's interesting to see the huge majority AMD has with members of this site.

That says something.

This is a site for PC enthusiasts, not (in most cases) people who buy a system from Dell because they have no idea how to build their own.

I have friends who hassle me daily to join the dark side and go Intel because they have the lead now and for the foreseeable future.

To me, it's not about the money. It's about building, learning, OC'ing and tweaking. I enjoy the feeling of accomplishment I get by taking what your average uneducated "I just ordered a kick-ass new PC that I saw on TV" Intel fan thinks is a substandard processor and OC'ing the chit out of it for next to nothing.

That's why I choose AMD....because I take pride in earning the performance my rig pumps out rather than just buying it.

Besides...what fun is there in OC’ing with a locked multiplier? LOL

Kgentatszu]

you can always increase the front side bus.....~LOL

KNIBBO

You just proved my point, but you don't get it, do you?

mike2h

that is cool that you enjoy tinkering, but me, i just want to make as few adjustments to the bios as possible, run that fsb up, & start playing!

__________________

mike2h

my computer is a p4 1500 on a asus p4t running at 1650 for over 18 months straight. my wifes is a abit be62 with a powerleap adapter running a 1200 tualatin at around 1450 for about a year now. never had a single burp or problem. extremely easy overclock on both. stock(retail) cooling all the way around. i gave up on amd back in the thunderbirds not just because of cpu issues, but mostly because of poor chipset support. imo. this appears to have changed considerably but i am stuck on intel now & asus p4p800 deluxe here i come!

Zardon

no offense, but have you ever actually tried to overclock an AMD cpu? ive been overclocking AMD for years, and I dont think ive ever seen this problem, unless of course you go outside the CPUs max frequency, this would of course in that case not be an exclusive problem of AMD.

WyreTheWolf

Firstly, lets look at two aspects... The PR value really upsets Intel users... so lets discuss that first:


Unfortunately, even some of the brightest people I know are subjugated to the megahertz myth… that more megahertz means more power. Intel established this as their method of rating processor speeds long ago, and as such, John Q. Public has fallen victim to these beliefs that the processor clock is all that counts due to Intel’s obvious dominance in the processor market. How then can an Athlon XP 1800 (1.53GHz) beat a P4 2GHz?

Glad you asked! In reality, there are many factors that contribute to a processors speed. For the sake of simplicity, we will ignore most of these such as architecture differences & development controls (SSE, MMX, 3Dnow!, etc…) and try to stay focused on the speed equation, which as you will see, quickly dissipates the thought of more megahertz alone means a faster processor.


The Speed Equation:

Cycles per Second (hertz) X Instructions per Cycle (flops) = Processor speed

There we have it, and as you can see, the clock speed of a processor, or the number of cycles it goes through in a second is only half of the equation. The other half is how much work it gets done in one clock cycle, or the instructions per cycle (IPC). Please note that cycles per second is the “clock” of a processor, since it is based on time, and is therefore measured is hertz, whereas the number of instruction completed per cycle are measured in flops (seriously).

Now with the P4, Intel doubled the pipeline that the PIII had, from 10 stages to 20. What that means is that each stage gets done sooner (therefore more cycles) but less work gets done at each stage (instructions per cycle). This allows Intel to ramp up the clock speed making the processor appear, to the average consumer, to be significantly faster, in comparison to other processors, than it really is. This is why a Celeron @ 1.2GHz could match or even beat a P4 @ 1.4GHz, or why an Athlon XP 1800 @ 1.53GHz can beat down a P4 @ 2 GHz. Both the Celeron (based on the PIII) and the Athlon have a higher IPC than the P4, meaning that the P4 has to have a large clock advantage just to be able to operate as fast as either processor.

A good analogy for this would be if you had two dump trucks, one is just the normal dump truck (the Pentium 4) and the other has a second trailer attached to it (the Athlon) Now if you need to haul a couple tons of dirt to another location several miles away, the trailer-less dump truck is capable of driving faster than the other truck since it doesn’t have to lug around the additional trailer, allowing it to make more trips in a given period of time (clock). However, the dump truck with the trailer can carry more dirt per trip than his comrade, and therefore doesn’t have to make as many trips. All my babbling equates to the fact that he P4 needs at least a 300+MHz gap just to break even with a P3 – and 400-500MHz in the case of the Athlon (An Athlon XP 1800 runs @ 1.53GHz and nearly sweeps the P4 @ 2GHz in every bench mark you throw at it) and it will continue to need even greater MHz advantages in the future because Athlons do more work for every extra hertz they are raised.

Please note however, neither way is wrong or right, just different. And each has its own set of problems. When you’re trying to raise the IPC, you have to figure out how to modify or design your architecture to allow for more work to be done. When raising the clock speed, issue of heat dissipation and power consumption become a thorn in your side along with packaging issues. An example of the later is that not until the G5 (set to run at 1-2GHz), which isn’t officially released yet, has any PowerPC processor required an active cooling solution. What was the last Intel chip you could run with merely a small heat sink and no fan reliably? In the Consumer segment, Intel’s dominance and influence give it an advantage however. The problem lies in the amount of knowledge Joe Schmoe off the street has, which sad to say, isn’t much. And he therefore believes that the P4 is faster solely because it’s a bigger number… oh, the woes of not being number 1!

For reference, Intel is the only processor manufacturer that actually uses this method of pumping the clock speed, although VIA appears to be following in its footsteps – or VIA’s processors are really just that bad. IBM & Motorola design/manufacture the PowerPC series of processors, the ones that power Macs (G3 –consumer segment, and G4/G5 –performance segment, respectively). And even though the clock speed of these processors are less than half of Intel’s, they run just as fast, and in some cases faster than Intel’s chips, as they have put their efforts into upping the IPC (Instructions Per Cycle) of their processors, with very low jumps in clock speed. For reference, AMD maintains a balance of the two options, preferring to up the IPC of its processors with moderate jumps in the clock speed.

Now, this was very, very brief, not even venturing into the server or mobile segments, or the detailed disadvantages of having that longer pipeline, among other things, which can further hinder performance. Nor do we discuss the architectural differences between the processors, aside from what was absolutely necessary. All I want you to come away with is to not judge a book by its clock or flop, this type of situation is all over the computer industry. The only advice I can offer is to know what you’re getting, and know how it compares to other product options. By: Chris Lockner




Secondly, Intel users keep quoting that Intel P4 uses less power, is cooler running, more compatable, and so on... Lets discuss this next:




In the race between AMD, and intel; look who keeps falling down:

Intel, hoping to steal AMD's Opteron thunder, days before the Opteron release, produce a great new workstation chipset.

But the Processor contains a flaw...

This is typical of Intel when faced with competition. They push to get products out early and fall down every time.

AMD had a 233mmx processor, Intel had a 200mmx processor. Intel's push to get the PII released on it's proprietary slot prevented them from finding the floating point bug.

AMD had the Athlon, 50mhz faster than the PIII, at a better price. Intel needed a fast, low priced version of the i820 chipset that used SDRAM instead of RDRAM. The Memory Translator Hub had a flaw and 3,000,000 motherboards were recalled.

AMD was pushing 1.2 GHZ, Intel over clocked the 1 Ghz PIII to 1.13 Ghz but had to recall them a month later. This was a beautiful example of a paper launch, because the recall affected less than 200 customers.

With the failure of the 1.13 Ghz PIII, The Pentium 4 was pushed to early release. The processor was available for 6 weeks before any motherboards were released. Although no buggy chipsets made it to market, the delay was widely reported to be due to a flaw in the chipset.

Competition with Nvidia's dual channel nForce2 for AMD forced Intel to produce a dual channel chipset of their own, Granite bay. Granite bay was released for the holiday rush as an interim solution while waiting for Canterwood. Granite bay has a flawed AGP implementation, but no recall was offered. If you want Dual Channel with AGP 8x, you buy an Nforce2 or wait until Canterwood.

Well, Canterwood is here as a high end workstation chipset along with the 3.0 Ghz 800mhz fsb Pentium 4, and it arrived in typical Intel style, flawed. Interestingly, despite Intel halting shipments to the channel, the channel is full of 3.0 Ghz Pentium 4s. What happens to the customers that jump on it now? Do they get a replacement or is this another case of Granite bay style buyer beware? It is interesting that Canterwood is positioned as a workstation chipset. A typical product release would cover the complete market range from budget to high end solutions. Especially considering that Springdale is the same chip sans memory bypass. In retrospect, June would have been a better launch date for both chipsets from a technical standpoint since it is certainly not ready now. Perhaps Serverworks' workstation chipset for Opteron got Intel a little spooked. Marketing calling the shots instead of the engineers?

But Intel machines are much more stable than AMD machines right? Obviously not. The only good Intel chipset in recent memory is the E7501, which apparently only served to upset their previously exclusive partner, Serverworks.

But they run cooler right?

68w maximum for a Barton cored Athlon. 74w average for the 3.06 Pentium 4

But Intel is compatible with more software, right?

Yeah, sure, just take a look at Intel's list of games incompatible with another one of Intel's buggy chipsets, The 845GL. Extreme Graphics. Yeah, extremely bad graphics!

But Intel has Hyperthreading!

1 x 3.06 Pentium 4 costs $468 2 x AthlonMP 2600s cost $426

Why buy one processor that is pretending to be two processors when you can get two actual processors for less money?

But Intel has SSE2. That exclusive advantage will be gone on the 22 of April.

But, but, but. No, No, No.

Intel's monopoly has proven to be quite brittle, cracking under the slightest competitive pressure. I do not trust their technical prowess when they are under pressure because it has become so obvious that their marketing department is making technical decisions. Processors and chipsets are not like software that can be released as a working beta and patched later. Systems established in the gaming world, where the publishers call the shots and the developers jump, do not work in chip fabrication. It has to be released when it's done, not when the marketing department decides they need to sting the competition.

All this is made worse by what Intel does to the market as a whole, when they are not doing well. Intel's missteps have had a tremendous impact on the industry, destroying consumer and investor confidence. At the end of 2000, Craig Barret warned "This was a year of record annual revenue and earnings yet, slowing economic conditions impacted fourth quarter growth and are causing near-term uncertainty." He was faced with AMD going from 10% market share to 34% market share in a year. Wall street took Barret's word, as gospel, that the entire market was in decline and not just Intel's market share. Intel is a market bellwether so the market did drop, just so Intel would not have to admit that their missteps and AMD's superb execution allowed AMD to grab market share. Nasty business. Intel has a sub-standard product and they are reckless with their power.

The market is now better, all things considered, having for the most part recovered from Craig's fraudulent statement. Everyone but AMD is posting profits again, that due mostly to Intel spending down their 10 Billion dollar cash reserve they acquired by lying to their customers and producing junk silicon. 10 Billion dollars buys plenty of predatory pricing. AMD will be fine. The industry has rallied, like it never has before, around Opteron. The future is bright with crazy fast, low priced computers, but only as long as we have two robust competitors in the processor segment. As much as I want Opteron and Athlon 64 to succeed, I would not want x86-64 to be a knockout punch for Intel's glass jaw. AMD might play nice for a while but it would not be long before the prices on their top processors climb into the thousand dollar plus range like Intel's were at the height of their monopoly. It is good to know that Intel has yamhill or *asdf waiting in the wings to insure that when x86-64 is in it's prime, Intel might have a more competent response than they have historically shown. By: Jason Pippin




I have ripped this info shamelessly ... because I could not have said it better. I have owned, and built more Intel systems than any of you could shake a stick at. I presently own AMD.

AMD 1600+ (1400MHz) Palomino @ 1845MHz EVERY SINGLE DAY, PERFECTLY STABLE. (for those Intel users out there that say that AMD cant overclock and remain stable)

It has nothing to do with quality... Intel has had some of the BIGGEST FOOK UPS IN HISTORY ! ! ! Yet, somehow Intel still retains an aura of QUALITY about them... Ladies, and Gents ... this is called BRAIN WASHING... Pure and simple.

Stop falling for all the hype, and marketing... Use your heads, stop calling names. Read, mature, and learn... Time to grow as a person, and take into account more than just the size of your ... errr.... clock!

killzone

Currently using a P4 2.0 gig on a soyo dragon ultra board but have used amd in the past and had very good results with them and price wise they are very good also. I dont over clock especially on amd because of the negative feedback Ive heard about over clocking amd. If I decided to do that I would definately go intel and Ill never run anything under a p4 again anyway. Im happy. Its not like I have 3 or 4 extra processors sitting around in case I fry one. Peace

__________________
"THE TRUTH IS OUT THERE"

WyreTheWolf

You didnt hear any negative feedback about overclocking AMD on this forum. AMD processors (especially the newer cores) are OVERCLOCKING BEASTS.

Kgentatszu]

wolf, from the above discussion, are you saying that celeron is actually faster than a pentium 4 processor??

WyreTheWolf

Maxwell, I think you need to reread those statements.......

Now with the P4, Intel doubled the pipeline that the PIII had, from 10 stages to 20. What that means is that each stage gets done sooner (therefore more cycles) but less work gets done at each stage (instructions per cycle). This allows Intel to ramp up the clock speed making the processor appear, to the average consumer, to be significantly faster, in comparison to other processors, than it really is. This is why a Celeron @ 1.2GHz could match or even beat a P4 @ 1.4GHz, or why an Athlon XP 1800 @ 1.53GHz can beat down a P4 @ 2 GHz. Both the Celeron (based on the PIII) and the Athlon have a higher IPC than the P4, meaning that the P4 has to have a large clock advantage just to be able to operate as fast as either processor.

KNIBBO

This is why a Celeron @ 1.2GHz could match or even beat a P4 @ 1.4GHz, or why an Athlon XP 1800 @ 1.53GHz can beat down a P4 @ 2 GHz. Both the Celeron (based on the PIII) and the Athlon have a higher IPC than the P4, meaning that the P4 has to have a large clock advantage just to be able to operate as fast as either processor.


I know you're not asking me, but I couldn't help myself. Look at the clock speeds and you'll understand.

WyreTheWolf

Yes look at the clock speeds again... 1530MHz vs. 2000MHz the Pentium4 has a 470MHz advantage in clock speed... but doesnt compare in real world tests.

And the Celeron that is mentioned is the one based on the PIII, not the "new" celeron based on the P4.

Kgentatszu]

[QUOTE]Originally posted by WyreTheWolf
[B]but doesnt compare in real world tests.



first of all, thankz for all the reply......sorry but i still dont understand (stupid!!!!)
doesnt compare in real world test??? the amd is faster, right??(i'm confused now)

WyreTheWolf

[QUOTE]Originally posted by Maxwell1234
[B]Let me put it this way... I have an AMD Athlon XP 1600+ Palomino that is running at 1845MHz It takes a 2.4 - 2.5GHz Pentium 4 to do the same work that my 1845MHz Athlon does.

BWX

Thanks for thr article Wolf, I have been reading that exact same thing for years- IPC DO matter- And Intel has had many problems with releasing products too fast that are flawed. But then you get people in here that will come out and say things like "AMD SUX" and "AMD have no quality"-- Thanks for explaining it to them, although some of them I imagin won't understand no matter how much effort you exert trying to explain it to them. You didn't even start with the business of how Intel crippled the Celeron by stripping it of cache. Some people will never learn, I guess I don't care if they do or not anyway- let them think what they want, and say what they want- they are only showing their true intelligence.

__________________

WyreTheWolf

Well, the way that I look at it... AMD, and Intel both do the job. I have shown that in no way does AMD "suck" (feel like I am taking care of a room full of Kindergardeners)

The AMD CPU is a quality component that gets the job done... The only real difference is: IT GETS IT DONE FOR FAR LESS GREEN. Some people are willing to fall for the Megahertz Myth... I am not. Neither is my checking account.

scobywhru

What about CISC Chips we are talking about RISC Chips here and bringing up IPC and all that but CPUs fall into 2 categories as I know them and its not Intel or AMD, but CISC or RISC chips. Also why is no one bringing up transmeta its not like they are Horrible for what they are look at The green mile super computer, built on transmeta chips if I am not mistaken. All this AMD vs Intel is a bunch of cow dung, tell your experience say your opinion and respect what others have to say.

__________________

BWX

I don't have to respect someone saying "AMD have no quality" or other nonsense when they don't know what they are talking about and neither does anyone else. Why aren't you using a transmeta chip in you desktop? That's the same reason why we are talking about AMD and Intel, because that is what 99.9999% of people have.

__________________

krazy1

Could not be said better.. AMD always!!!

__________________

JCLW

My dual PIIIs still have enough kick for everything I do.

- JW

No_Style

But I got Intel once more

Why? A 2.4B --> 3.33GHz? Besides the fact that it topples almost anything the market can throw at it (Stock) and aircooled. As for the price/performance ratio. I do agree...If it were at stock and minor O/Cing was involved. If anybody came up to me and asked my opinion on which processor to get and they were not O/Cing..I'd throw an AMD chip their way. Low end to mid range wise, their price/performance ratio is kick ass. But right now: 2.4C with HT and 800 FSB. I don't know about my previous statement anymore....

The gap between AMD and Intel's High End is widening as well. Just read the reviews ppl

__________________

Kgentatszu]

well, at least that clear things up a bit, thankz for the note

satech2001

AMD AMD AMD!!!!!!!

Sorry...i had to get it outa my system...

Damnit Wyre...you gave me a headache reading that thing... ...not that you havent told me a zillion times allready...but i really really enjoy reading that thing

__________________
[COLOR=red]Intel P4 2.4 @ 3.544 mhz (1.75v)(12x295) [/COLOR] [COLOR=orange]Aircooled with Thermalright Slk-800u+enermax 80mm(3200 rpm)[/COLOR]
[COLOR=red]Abit IC7 Updated nb cooler (1.6 bios)[/COLOR] [COLOR=orange]TwinMos 2x256 dual channel @ 236 fsb(2.8v) 7-3-3-2.5[/COLOR]
[COLOR=red]Club3D Radeon 9700pro @ 380/338[/COLOR] [COLOR=orange]Maxtor diamondmax9 8mb cache 80gig[/COLOR] [COLOR=red]Case Y-Y cube highly modded[/COLOR] [COLOR=blue]best marks so far: 3dm2k3-5994 3dm2k1 19886[/COLOR]
My modded case
3Dmark2001se

FuNsTeR

i had a a p3 600 mhz coppermine , lets just say after that cpu i will never buy intel ever again first one maxed out after 2 weeks and no i did not O/C it 2nd one lasted 2 months before it started to act funny the 3rd lasted 4 months before it burnt out and neither were oc'd , at first i thought it was my Mobo ... but i got it checked with a freind who ran his own pc shop and he said my Abit BE6 rv 2 BX440 mobo was working fine and the problem must have been a dodgy batch of coppermines and intel snobs have a cheek to slag reliability problems concerning Amd products

ever since its been Amd all the way 1200 Athlon no probs with the processor worked a treat now my mate has it in his pc , currently have a XP1800 runs like a dream next Cpu will also be a Amd processor

__________________
"They say when you play a Microsoft CD backwards you can hear satanic messages ..... but that's nothing, if you play it forward it will install Windows"

No_Style

CPU Speeds. First thing I tell people when they ask me about AMD is this: "Forget MHz, it's performance benches that matter. Just like pure Horsepower doesn't mean anything" I've personally never used AMD, probably never will. I trust Intel products and every single Intel product I've used continues to amaze me. I've always given people choices when they come to me looking for PC Building advice. But with the current Intel C series. I can't stress enough how feature packed that baby is

__________________

mike2h

a lot of the things you say are very true. two things you got wrong though are dual channel - intel's 850 has been out for at least 2 years, & you can get hyperthreading for $175(with 800 bus)
outside of heat issues, wich is not the problem it used to be, the biggest problems with amd has been their lack of top quality chip support. amd, in the past, has had little or no control over the quality of the chipsets made for their cpu's. the k5, k62 & 3, athlon & first tbirds had very poor chip support. in fact it was so poor that amd was forced to make their own chipsets . which were good but had the downfall of being less available & a little pricier. some of the chipset issues carried through until very recentley. intel has had their share of problems but they are good at pr'ing them into non existence and/or fixing the problem very speedily. they have more money, better bigger production & tighter control over non intel chipsets. & again more money. this does translate into better real world overall quality. now do not get me wrong, i am very thanful that amd is around & makes cpu's as good as they do. for 3 reasons- 1) without amd we would be paying much higher prices for less performing cpus. 2) amd IS geat bang for the buck. 3) choice! choice is great, also lets us have these debates!

BWX

Well actually the K6-2 was pretty stable- at least the one my girlfriend had in her old computer in 1997- it still is sitting in the closet- It actually has an Asus P5A mb that was also stable as hell- Even back then the thing never acted funny or froze once-- I ran that pc for a while before I built this one- no problems at all- And anyway, if that is your argument for why you think Intel is better then you need to forget about it- that is in the very distant past- let's talk about more recent chips, like the T-bred B's?

And hey, didn't you READ what Wyre was writing? Didn't you hear about the premature release of all the Intel products- (talk about quality?) didn't you hear about the huge recalls and problems w/ many, many intel products over the past years? And even not so long ago. It's like people who think intel is "better" just seem to overlook all those things. Not to mention Intel's marketing which is questionable at best.

If I had more money than brains I MIGHT go but the best Intel has to offer just because it MIGHT give me a LITTLE more performance than AMD's top chip- but those are big maybe's- especially when you start talking about overclocking potentials with normal air cooling.

__________________