I suggest the use of a processor with a 2nd level cache of >=1MB. That should be ok for our "bros java programmer" and he should make sure the processor and the RAM clock the same freq. (I mean front side bus and ram clock speed should be same)

nice one. the more, the merrier!

*sigh*

ok. I decided not to be more explicit previously, but I'll interpret the responses from naijafan, romulus and christino as requests for such.

I'll just request that u guys take a seat, relax, and let me do a little talking.gonna be a long post. Not trying to put anyone down here, hope we all understand! And if anyone has an issue, lets just put it up and settle it like men, so this thread becomes an informative one to anyone who comes here wanting to know thw answer to "AMD vs Intel: Which Is Better?"


aaight, first Romulus.

1) it is not a given that a 64bit chip will outperform a 32bit chip
2) with CPUs, speed and performance are the same thing, so one company can't be after speed while the other is after performance.

explanations:

1) When you call a chip an 'x' bit chip, you are generally talking about word length. When word lenghts were still one or two bytes long (8 and 16-bit chips), it made a HUGE difference when you moved from one to another, because all of a sudden a task that took several instructions suddenly took a few. you didnt have to muck about with lsb and msb as much as usual (u sound like a comp.sci/ee person so i guess u know what that means, ) and so your speed increased dramatically.

Then we go to a default word length of Dword, which is what 32-bit chips have. Suddenly your integer range is from 0 to close to 5 billion. Not many people within the common problem domain need that kind of precision very frequently, so suddenly the operations that get a speed boost are not as common. Nonetheless, 32-bit chips gain prominence around the time of RISC development and many of the advantages of RISC procs (which btw are not actually due to risc itself but thats another story) are attributed to 32-bit procs. Similarly, the ablity to natively adress up to 4 gigs of RAM without mucking about with segments and offsets (pleeeease im begging with tears in my eyes, dont get bored, just listen, ) meant we had to deal with stuiff that previosuly was not feasible on chips with less adressing power- so we add stuff like protected mode, pre-emptive multitasking etc and the mass media trumpets all this as '32-bit features' when, even though these are xtixs of 32bit chips, have nothing to do with the word length.

fastforward to 1996. The 64-bit brigade comes in trumpeting, and all of a sudden it is glaringly apparent that there is little you can add to a chip that requires a higher word length. Oh there are high precision applications, but specialised chips are available to handle such.

for a decade we use 32-bit chips on our desktops.

All of a sudden , today a use has emerged for 32-bit chips:addressing memory above 4 gigs. All of a sudden 64-bit is the new buzzword. Intel tries to do the 'old thing' and use it as an excuse to totally re-invent the processors design. AMD just puts out the same old processors with double-sized registers, 64-bit extensions to the instruction set, and a longer word length and slaps on the name 'athlon 64'.

Guess who succeeded?

in short: word length is not only overrated, you are probably not using it, and if you decide to go thru the wahala of reinstalling windows and all your drivers just to be '64-bit compliant', goodluck telling the difference.




2)

You say that since an amd has a clock speed less than the pri number, their chips are slower, and somehow have better 'performance' due to 'pipeline bursts' and 'ppc' or whatever.

unfortunately you fail to define what you mean by 'speed'. So let me help out.

processors consist of multiple units, the ones that concern us being the alu and control unit. the ALU is the 'powerhouse' where the work is done, and the control unit is the bit that makes it all happen in the realm of reality, by imposing real-world controls on the electron flow that cause the logic to actually make sense. In short, tha ALU does not function without the control unit. The control unit tells registers when to shift, pushes insructions down the pipeline. The work is done th teh arithmetic/logical unit (ALU)

now a major component of the control unit is the clock. The clock is an oscillator that ticks so many times a second, with each clock tick creating a burst of energy that moves the processor forward one 'state'. The reason the speed of the clock has always been associated with processor speed is:

based on chip design, if you run a processor clock too fast you will destroy the processor. SO a better processor accomodates a higher clock. So a faster clock means more performance!!!!


however if we take this out of context we are commiting a DANGEROUS BLUNDER, one that intel has succesfully capitalised on for at least 3 years of marketing underperforming processors. ALlow me to explain.

now when we say clock speed, we mean exactly that: CLOCK SPEED and NOT processor speed. we are talking about the speed of the clock. If i take a chip and attach a control unit with a higher clock speed, that does not fry the electronics of the chip, i will get a proportionately better performing chip! However, if i simply replace the ALU of the chip with a better, more efficient one, I also get a better performing chip!

we are measurng the performance on WORK DONE PER UNIT TIME. in short, SPEED.

get that? what you call 'performance' or 'muscle' is nothing more than SPEED.

and what you call SPEED is nothing more than CLOCK SPEED, which is supposed to be a VARIABLE in the calculatiuon of actual SPEED!!!!!!!

so!

to rewrite your statement:

original: Intel is after speed and AMD is after peformance

correct: Intel is after CLOCK SPEED and AMD is after ACTUAL SPEED

which shows that, at the end of the day, intel is after a BIGGER NUMBER THEY CAN PUT ON THE BOX TO MISLEAD THOSE WHO CANNOT OTHERWISE TELL WHAT SPEED THEIR PROCESSOR IS AT , while amd is after A HIGHER SPEED EVEN IF IT MEANS THEY HAVE TO INVENT A NUMBER TO PUT ON THE BOX.

amd's prc numbers are nothing more than customer information saying 'we are as fast as an intel proessor clocked at this speed' so that customers are not misled!

pipeline bursts, ppc and so on are all measures of SPEED!!!! we are ALWAYS talking about SPEED! there is nop disconnect between SPEED and PERFORMANCE! CLOCK SPEED is just a VARIABLE used in finally determining the EFFECTIVE SPEED! clock speed is the speed at which the control unit orders things along! it has nothing to do with the proc except for the idea of chip design acommodating higher speeds! MY P4 3.0GHZ WAS DESIGNED TO ACCOMODATE 3000 MILLION CLOCK TICKS PER SECOND AT THE EXPENSE OF ACTUAL PROCESSING SPEED! INTEL SACRIFICED VALUE FOR A NUMBER THAT I COULD USE TO BRAG TO MY AMD-OWNING FRIENDS!

my p4 uses deep pipelines to accomodate a higher clock speed. any compsci person in the house knows how DAFT this decision is. the longer your pipeline, the worse your penalty for branch misprediction or any other thing that could cause a pipeline flush. stupid? well it helps to bamboozle customers!

this brings me to the next dude,

NAIJAFAN

yes, I did say ordinary users.

remember what i was saying to romulus just now about customer bamboozling? well, it works when the difference is so small that only gamers (eyes fixed on the fps meter) or video encoding people (looking at encode times) ever notice the difference. Well, recently, intel dumped the netburst (pentium 4) architecture and fired a whole load of top-level staff responsible for it. what could prompt such, pray tell? simple.

people notice 2 things about prescott-class (read: modern day) pentiums:

they are hot.
they hang.

Aha, you say. Hanging doesnt mean slowness. It means instability. They should keep the procs cool. Except that, this 'hanging' is how an 'ordinary user' would describe what we compsi types would call 'blocking conditions'

computers nowadays are so fast that frequently, the hardware is able to do stuff at a speed where it makes itself available to a process but is not released by the previous one because it is still 'computing' and needs some time to finish. this is in contrast to the old days when we were always adding RAM and disk space since those were the bottlenecks. nowadays, most of us talk about 'cache', because we already have too MUCH ram and disk space. We now get to ACTUALLY SEE a process working, while the rest of the computer goes idle. And this, my dear friend, is what a user means when he says 'its hanging'.

he means: explorer is blocking. taskbar is blocking. WORD is blocking. IE is blocking. Printer drivers are blocking.

Now I work for a Bank that has deployed over a thousand machines in head office alone. I write code and have to suport it at the same time (stupid i know, tell my bosses). Most of the time I get calls about 'hanging' and people BEG me to install on the 'black pcs' or if they read the cases,

'the athlons'.

why?

they dont 'hang' as much!

translation: they are faster!

and the reason for this is that netburst enabled amd to gain a ridiculous advantage in pure speed, hoping that no-one would notice. Unfortunately, this did not turn out to be the case!



Oh, and finally I meant it: benchmark tests lie. And im not talking about nvidia tweaking drivers or whatnot. Im saying that such things as 'super pi' or whetstone calculation tests do just that- test how fast you can calculate flops or super pi. Finish. The ultimate benchmark test is just to USE the damn computer. And when one is noticeably snnappier than the other, aaaaaaaha.

my point: in the past, no-one cared about the CPU cos it was memory, vidcard, etc you looked at. But TODAY, cpus suddenly matter. because people realised AMD is faster.




finally,

CHRISTINO

its simple why your bros comp is faster. BOTH your cpus are capable of running the games smoothly. however BOTH your vidcards are most definitely NOT. his vidcard is obviously better than yours, making your cpu advantage moot, cos your extra CPU power is useless (since you are both GPU-limited and not CPU-limited).

cpu-limitation only occurs when both test systems are maxed out on gfx power. At that point slowdowns due to ai, etc will determine the winner.


I use an ati radeon X800XT 256mb GDDR3. prior to that I used an ATI RADEON 9800 PRO 128MB. no matter what the speed of whoever's laptop came to my 'battle chamber' was, or whoever's gaming pc i got to play on, my p4 2.26 GHZ beat them all mercilessly due to the superpowered vidcard. Only when i began to do cpu-limited stuff (dreamcast emulation, video compression for my PSP) did the inadequacy of my main proc become apparent.



conclusion: As you rightly poiint out, for many applications, the cpu-type is not important, because they will eprform identically and the deision will be left to peripherals (in this case vidcards). HOWEVER this is a thread about which proc is better, implying that WE WANT TO DO SOMETHING CPU LIMITED! running a risk asset schedule computetion program is one, which is why my users kow the differenct between processors. Playing emulated games is another, which is why me (and many in my circle of friends) will laugh at any person that yans about intel being faster. (disclaimer: c2d has defeared AMD. but I accounted for that earlier and explained why i dont advise buying it. PLUS c2d's are low-clock processors)



So there we are. a whole load of ranting. PS absolutely NO PART of the above post is useless grammar intended to confuse the reader. IT IS ALL RELEVANT so please read and shoot down any parts you feel need shooting, but please indicate the bits you agree with in order to keep discussion civil and short. If any bit is unclear its my fault and i will be glad to explain further. Thanks to all.

i was reluctant 2 use AMD wen i knew abt them 3 yrs ago.The likes of AMD K6's are as old as pentium 11's or Cyrix6x86's.i ave not really seen much diference in both since AMD's are high end processors and they zip reasonably well at image editing or gaming,i'll be dumb 2 burn alot of money on intel wen i can buy chips of intel speeds so cheap, Thnx to AMD

Word darkman, word,

But have you heard of 3DMark05 I understand there's even 06 now, maybe if you can try that on your system, the download size is about 300MB or more. I have the full version of 05 if u need it, but i'm in Lagos. Google 3dMark05 and read up the reviews on it. I understand www.guru3d.com even have a forum where users come in to drop their views and their benchmarks on VGA's processors, mainboards and stuffs.

For Benchmark results, check http://forums.guru3d.com//showthread.php?t=179987

or check www.hardwaresecrets.com

cheers. FYI, Athlon 64 FX's now 218 dolls.

None of Them

Pls,wotz d price range of an AMD Processor?

dakmanzero:


1. they don't 'hang' as much!

translation: they are faster!

,

2. Oh, and finally I meant it: benchmark tests lie. And im not talking about nvidia tweaking drivers or whatnot. I'm saying that such things as 'super pi' or whetstone calculation tests do just that- test how fast you can calculate flops or super pi. Finish. The ultimate benchmark test is just to USE the damn computer. And when one is noticeably snnappier than the other, aaaaaaaha.

3. PLUS c2d's are low-clock processors

Let me deal with just these three one after the other briefly.

1. Please citation needed
2. You need to visit overclock.net! Its a forum right. And also, citation needed
3. Its dual core for god's sake! Both cores running at that speed!

1luv.

@ NaijaFan,

Share you IT adventures here: https://www.nairaland.com/nigeria/topic-26409.0.html

naijafan.

Asking for citations for point 1 is ridiculous for obvious reasons. What do you need, an interview?!! jesus. It is an informal description of a daily occurence, and the only thing backing it is my credibility. If you choose to believe I pulled that whole story out my arse for no reason whatsoever than to oppose your point of view, you are entitled to that opinion. Sheesh.

You do NOT need a citation for point 2. It is not a statement of fact based on reference, it is merely a statement of fact based on pure common sense. A benchmark measures WHAT IT MEASURES and nothing more. You dont need a citation to figure that out. The extrapolation required to translate a benchmark test into a reasonable estimate of performance is highly subjective and problem-dependent, but uit is easyu to bamboozle the general public with big numbers, so hardware vendors just post them anyway and leave you to jump to your own wrong and inevitably 'seeded' conclusions.

I do not need to cite anything for point 3. just head to intel.com and check the advertised clock speeds for c2d's. And, if you say that you need 'proof' or a 'citation' that two cores running at a certain clock speed is in no way similar to one core running at twice its clock then I have nothing more to say.

AMD vs intel, linux vs windows, automatic vs manual, playstation vs xbox, good vs evil and so will it continue for ever

@darkmanzero

maybe you didnt read my third point well, i didnt request for citation.well, that's what happens when you're out to attack. you miss even the most basic elements of a post.

you pointed out to christino that you wouldnt advise anyone to buy a core 2 duo because they are new and have low clock speeds. i simply replied by saying they are dual core for crying out loud! hope that makes you understand.

now coming to the others.

no idiot will translate a hanging processor to mean a slow processor. that's why i requested for your sources. i have a feeling you just want to write seemingly big words so people would appreciate your posts. nice one there!

now to the benchmark stories. have you tried benchmarking yourself? have you visited overclock.net? if intel pays pc world and others, why doesnt amd do so as well? how come the big numbers mostly turn out to be in favour of intel? as long as you continue to argue blindly (or maybe not pointing us to your sources) i would rightly continue to request for 'em.

do you know the first gigahertz processor was from AMD? and do you know the pentium III 933Mhz outperformed it? you see, the processor race is a really competitive one. and as i have said earlier, neither intel nor amd stays a winner for too long.

Hmmm.

I was going to aologize for the point 3 citation thing, but I guess that will need to wait, cos you comitted the exact same blunder yourself, if your explanation is to be taken seriously.

I did NOIT tell cristino not to go c2d because of low closck speeds. for the sake of osutompolor's kunu-filled beard I am the person here that has been arguing in FAVOUR of lower clock speeds, for Jesus' sake. If you BOTHER to read my posts AT ALL you will see that i acknowledge that the c2d beats AMD at every turn!

my points are:

1) before c2d, Intel has been clearly behind AMD. this is so glaringly obvious and such common knowledge that I cant imagine why the hell I am even bothering to give this issue time of day! For GOD'S sake are you ttrying to tell me there was a single prescott processor that even smelt AMD's rear lights in terms of performance? If you are then all I can say is, oh dear.

2) the c2d is intel's AMD-killer, AND IT DOES SO WITH A LOWER CLOCK SPEED, THUS PROVING MY POINTS!!!!

3) I say christino should NOT go for c2d's not because of low clock speeds BUT BECAUSE THEY ARE STILL SHINING NEW AND THUS OVERPRICED! Look, we all live in Naija. Maybe u live in jand or yankee where its cool to get yesterday's processor NOW. In naija that dont work. No tech support, no compatible hardware, no nothing. He's better off waiting a year for ikeja to become c2d compliant

BUT I ALREADY SAID THIS. READ MY FR!CKIN POSTS.



As for the other issues you raised,


(a) the hanging=slow thing was tjust to make you see that the difference between amd and intel had become so great that even end-users could see it. of course any blithering retarded mongoloid slowpoke knows that hanging doesnt EQUAL slow. HOWEVER THIS IS HOW THE USERS WILL COMMUICATE THEIR EXPERIENCES. Jesus.

The hanging thing is by no means p[roof thyat amd has been consistently killing prescotts for over a year. THAT is pure common sense AND common knowledge.

(b)
for jesus sake read my posts and stop jumping to conclusions based on a cursory scan of the contents. i say benchmark tests lie not because anyone is paying anyone off. I say benchmark tests lie because they, aaagh read my posts. I dont have to keep repeating myself. Ive said this exact thing already. Read the old post and respond to that, if you take exception with it.


finally your quip about first gigahertz processor coming from AMD is utter and complete bullshit. Please edit it out of your post before it discredits you, because most of what you have written, even if it conflicts with my views, is quite insightful and well-supported (i imagine that that is your main issue with me, I dont support my facts as well as u do) but THIS ALONE is pure bullshit. Go read up on the AMD thunderbirds, and on the twilight days of the p3, including the lil tidbits on the legendary tualatin processors. Much wisdon awaits you there.

wow. maybe my 'quip' wasnt explanatory enuf. amd's crossing of the gigahertz line was forced. a processor who's clock originally was 800MHz was overclocked and a special cooling system was devised for it. hell, i read the reviews back then even in 'the economist' or so.

please i'll advise you not to contradict yourself anymore. whenever you get a reply from me, dont move your mouse hastily to the reply button. and if you do, take your time to read whatever you've said earlier to warrant my reply before babbling.

dakmanzero:

(disclaimer: c2d has defeared AMD. but I accounted for that earlier and explained why i don't advise buying it. PLUS c2d's are low-clock processors)

now can you see you said it yourself (except if the word plus has a new meaning)? of course i new all along you were advocating for 'low clock speeds' in favour of AMD. maybe i should have pointed that out to you before replying with my 3-point post. pardon me please. but i've still got a line for you. READ YOUR FR!CKIN POSTS.

we're supposed to be technical guys here. you vividly translated 'they dont hang as much' to mean 'they are faster'. for god's sake, what do you intend teaching the over 700 people who have read this thread?

and yeah, you said benchmark tests lie because they measure what they measure! right?! if the results do not xplain what you need your system for, then the processor is obviously not for you. and can you remember telling christino you wouldnt go to guru3d.com cos its paid marketing shit? READ YOUR FR!CKIN POSTS.

man! what do you mean by being c2d compliant that aint technical at all!

@darkmanzero

about the discrediting shit. i just stumbled on an archived page dated March 6, 2000 that says the same thing. http://www.dwightsilverman.com/cpq-gig.htm

just in case you're lazy to go there, let me drop a few lines

'Both Compaq Computer Corp. and Gateway on Monday said they will sell computers that use Advanced Micro Devices' new 1-gigahertz processor, but those who've tested systems say some users may not see a dramatic performance boost.'

'With Monday's announcement, AMD becomes the first maker of personal computer processors to sell a chip that runs at a gigahertz, the equivalent of 1,000 megahertz. '

'In doing so, Sunnyvale, Calif.-based AMD beat out chip giant Intel Corp. for bragging rights as the first to deliver a 1-GHz processor. Intel is scheduled to announce a gigahertz version of its Pentium III chip on Wednesday.'

'Michael Miller, editor-in-chief at PC Magazine, said a business-application test showed the Compaq and Gateway Athlon-based PCs were not as fast as a Dell Dimension with an 800-MHz Pentium III computer. Miller said the processor in the Dell has a built-in L2 cache.'

can u see that all ur babbling about all amd processors performing more work per clock cycle compared to intel ones arent fully true? man, i've got a brilliant memory. i read that six years ago in magazines. as said earlier, i have been into processors since the 90's. so calling my posts insightful is a bit of an understatement. they are facts.

and i still repeat. the processor race is very competitive. neither intel nor amd stays a winner for too long

touche,

I still take exception to quite a number of your points, and the hanging thing still needs to be explained, but my carelessness in dismissing your fact about the 1ghz thing appears to have shot my credibility to pieces.

Interestingly, I based my confidence on a memory of an article from back then, talking about the first thunderbirds (950mhz) and AMD's claims that they would outperform the 1ghz plus pentium4's.

Another article on emulators.com (rather famous, u prolly read it) also dissected the p3 and p4 and went to great lengths to explain intel's reluctance to allow p3's cross the 1ghz barrier, lest they are shown to be blatantly superior to comparatively clocked p4's- manifested in their handling of tualatin. The article gave the impression that AMD had not crossed 1ghz at that time. I suppose reading it again will manifest that as a false assumption,

Ah well. I suppose for the purposes of this thread I should concede to you since you obviously have a more stable basis for your arguments. Nonetheless, you may wish to do a bit of research on the athlon64 to determine the origin of my position.

thought this would go on forever. nevermind, we all learn everyday. i'll still kip up with my researches tho. tried buzzing u on zcl01 and got no response. when u've got the time, i'm on xolubi (Yahoo). hope to hear from you soon!

well i'll stick with AMD cos it got it all. AMD pros have always outrun Intels and it also goes for a cheaper cost. u can ckeck out their new chips barcelona and Phenom. it's the bomb!!!

My processor is AMD 64 athlon, and you know what? I love it.

`

Whoopee! So What! Why comparing? I agree w/adewaleafolabi's bottom quote.

adewaleafolabi:

AMD vs intel, linux vs windows, automatic vs manual, playstation vs xbox, good vs evil and so will it continue for ever

well, just to add a little gas to this. I saw this thread a few days ago and decided to ask at a local electornics giant. what i was told:

The biggest carriers of AMD laptops as I know it is slowly getting out of AMD processors because AMD processors do not have a cooling fan of which Intels have. AMD processors tend to have mor issues as the computer ages.

I have had an intel pentium laptop since 2003 and it is still going till today. I do not turn it off, it runs constantly.

Because it is cheaper sometimes does not mean it can wihstand the power surges in Nigeria, think long and hard about it. I am not picking any sides but that cooling fan for your processor is important.

Thanks!

i was gona create a topic on this, its a good thing i searched 1st!

i had a chat with dell consultants and asked them the diff btw the intel and amd processors! they told me its d same same performance and just diff brands!

after reading through this thread! and a cnet review, conclusion is AMD processors are powerful than the intel!

i would like to know wich AMD processor matches the intel core2 duo!??

anyone??

cheap- amd
expensive- intel

personally, i prefer intel becos they are much much faster than AMD. problem is that intels line are not yet optimized for todays sofware

AMD processor is just a chip







Intel processors are PROCESSORS



I'm Out

cut and paste