Will Athlon XP beat the Northwood P4 ???!

[mr_gobbledegook]

Hi cheery fruity chums,

Do you reckon Athlon XP (1.5Mhz) will be able to beat the crap out of P4 2Ghz and up and coming Northwood 2.2GHz P4 ??

I think together with NForce it will STILL be slower than Northwood because I reckon the increased performance of the Athlon 4s has been hyped. We will only see a real world performance increase of around 10%.

Only a 1.7Ghz Athlon XP will be able to beat the current P4.

I also think AMD have got a serious heat problem even with Palomino. They need to switch to .13 micron pretty soon if they want to stop thier chips frying.

<font color=red>UPDATED
You are forgetting Intel will be launching Northwood (.13 micron) close to the end of this year / early next year. Therefore AMD WILL HAVE TO compete with thier Athlon XP processor (.18 micron). I doubt we will see desktop Thoroughbreds in the shops until May 2002 the EARLIEST.
Until then Intel will have the upper hand, I wouldn't be suprised if they have a 2.6-2.8GHz Northwood P4 by then.

Since processor sales are down you may think high speed performance doesn't matter...HOWEVER now is the time AMD has to build up reputation as No 1. in performance so when the market does turn around people will go out and buy AMD processors.

The real test for AMD will be with Sledge/Claw Hammer (Q4 2002) It will be the make or break for them depending on its success. I am extremely interested in the disclosure they will be giving on 15th Oct 01, I hope they will give some indication of the performance we might expect from this unknown X64 processor.</font color=red>

<font color=purple> **I eat confusion and spit it out slow**</font color=purple><P ID="edit"><FONT SIZE=-1><EM>Edited by mr_gobbledegook on 09/24/01 01:21 PM.</EM></FONT></P>

 

[zengeos]

Weeelll...since P4 2GHZ system is about 8% faster than Athlon T-Bird...and since Pally should provide 10% performance boos as you yourself mentioned, Athlon 1.4 should best the P4 2GHZ.

1.5 should prettymuch be on par with P4 2.2ghz.

This is without using nForce. If nForce is as good as some previews say, Palomino 1.5Ghz coupled with nForce should outperform P4 2.2ghz.

Northwood is .13 micron.. as is AMD's chip AFTER Palomino, so let's see what performance improvements each company brings to bear with their respective .13 chips.

Mark-

When all else fails, throw your computer out the window!!!

 

[Boondock_Saint]

10% more speed per clock cycle is a HUGE preformace boost, relatively speaking.

Like he brought up, a Tbird 1.4GHz on par to a little slower than a PIV 2GHz. Now, that is the Willamette core, but putting the Palamino against the Northwood will have to be seen. I am putting my bet on the Palamino from a pure speed/price value compared to the PIV Northwood.

 

[flamethrower205]

10% certainly is a performance boost. However, with Northwood's Hyperthreading (better FPU), pally may be having a hard time.

Your brain: PC
You brain on drugs: Mac

 

[Guest]

I am using a Palomino 1.2 running at 1.4Ghz on a MSI K7 Master. Most benches I run (10x140mhz 8-8-5-2-3-2-2) show a small gap (in my favor) between my system and a P4 2Gig. Some numbers are waaay on the P4 side - SSE2 enabled, and some numbers are waaay, waaaay on the Palominio side - FPU x87 stuff.

With new MOBO BIOSes and the OGPGA coming with the new AthlonXP - not to mention nForce - I'd bet on the Athlon.

(BTW, the palominio gains much more from SSE than the P3 or P4 does, why? I don't know, something about the number of cycles per SSE instruction being less in the Pally than in the P3,4 - JC has the skivvy on that...)

Pally 1.2@1.4 (10x140)
MSI K7 Master
Swiftech MC462
Crucial 256 DDR x1

 

[AMD_Man]

I'd say that there will be no clear winner. The Athlon XP will excel in somethings while the Northwood will win in others

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor

 

[lhgpoobaa]

Dont forget that the all important comparison of the Athlon 1.4 and p4 2.0 was done NOT using either the nforce or the KT266A.
using those it would be safe to assume that 1.4=2.0 and 1.5=2.2.

cant wait till the DDR chipsets really mature with native PC2700 (333mhz) support

Religious wars are 2 groups of people fighting over who has the best imaginary friend.

 

[somerandomguy]

Will Northwood have a 133Mhz FSB (as opposed to the 100Mhz FSB it has now)?
Even if it doesn't, I think it will still outperform the Athlon XP, but not by a large margin.
The Athlon XP will, without doubt, be better in terms of price per performance.

"Ignorance is bliss, but I tend to get screwed over."

 

[Hit]

I agree with AMD_Man. There will be no clear winner and not just because the Athlon XP and Northowood will perform on par in different ways.

With the current economy there is little demand for these screaming processors. With e-comers taking such a long time to get up and running the demand will remain low. With DSL/Cable modem subscriptions still so high, people have little need to upgrade. Of course, e-comers and broadband have little to do with hardware however, they provide content people may want.

My point now being that as long as demand is low, I feel we will see Intel and AMD products stay at the same performance level. When demand improves each company will have the momentum to pursue market share and differentiate their products.

AMD seems to be positioning themselves by associating their product with the release of Windows XP. Intel just pumps up the GHz. Content moves these processors and I think AMD has finally figured that out. Hey, it worked of WinFast. Intel has not taken control of the internet like they had hoped with "Net Burst". So it's left to the consumer, the winner will be the one people buy.

 

[zengeos]

Palomino may have some difficulty in besting Northwood at the 1.4 to 2.0 ghz speed comparison. However, Thoroughbred should bring more performance improvements to the Athlon.

Northwood and Thoroughbred are both .13micron cores so the comparison between them should be more valid than a .18 to .13 comparison (.13 theoretically should see a nice performance boost just because of the smaller die)

Mark-

When all else fails, throw your computer out the window!!!

 

[Hit]

Next time I feel I should preview my post. e-comers HA!! that should be e-commerce...

 

[lhgpoobaa]

yes. i believe the 133 (533 quad pumped) bus speed is comming soon for the P4. before or during the nothwood release i dont know... PC1066 RDRAM is also comming soon to go with the 133 bus speed increase.



Religious wars are 2 groups of people fighting over who has the best imaginary friend.

 

[somerandomguy]

A 133 (533 quad pumped) bus speed at the Northwood’s release would be very impressive. It might even make me consider buying a P4 over an Athlon XP for my next upgrade.
It's a moot point I suppose; I don't have any money anyway.

"Ignorance is bliss, but I tend to get screwed over."

 

[Matisaro]

If you dont have any money, or have little money, amd is 99.9% of the time the clear way to go.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

 

[charliec2uk]

This is getting tiresome. The K7 CANNOT, WILLNOT, it as it is be able to use the kind of bandwidth offered by the nforce chipset. Go to anandtech and see the reviewed. There is not much special about the nforce for the Athlon.

For the future. As a applications demand more and more bandwidth, if the p4 's 4x bus is increased to 133 Mhz, so around 530 Mhz (effecticely) , this will feed the processor with 4.2 Gb/s. Thus I think the Northwood, wwill increase it's lead.
As for sledgehammer, compared to Mckiney, don't make me laugh. McKiney is going to open a six-pack of whoop-a$$ on sledgehammer, gameover. I think that AMD can stay in the competetion but only if they innovate. I been the very concept of sledgehammer is flawed. It was annouced very soon after the marketing for Itanium started. Knee jerk alert or what. (Please no flame mail I am just trying to voice resonable, probable theories, I own no Intel stock, (yet) might be a good time to buy though.

 

[Boondock_Saint]

"resonable, probable theories"

What a laugh. Your whole post was obviously Intel biased. Do you even know much about the Sledgehammer? Do you also realize that even though the bandwith will be increased, not much more of it will be used? Do you realize that DDR 333MHz is coming out? Don't you know that the Palamino and Thoroughbred are almost out or right around the cornor?

Please, read up on all the facts first. You are trying to compare a next-generation Intel chip, WAY off on the horizons, with what AMD currently has out. If you want to TRY to compare, compare apples with apples, not this garbage you just spouted.

 

[Raystonn]

"the Palamino and Thoroughbred are almost out or right around the cornor"

When are each of them scheduled to be released according to the latest roadmaps?


"You are trying to compare a next-generation Intel chip, WAY off on the horizons, with what AMD currently has out"

Which Intel chip are you speaking of here? Northwood will be released in approximately 2 months. McKinley will be released at the beginning of Q2 of next year. Neither of these is really "WAY off on the horizon." Would you agree?

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[charliec2uk]

Okay, deep breath, count to ten...

I shall address your accusations one by one:-

I think I know as much about sledgehammer as has been released to the general public. Essentilly it is targeted at the same market as Itanium, touted better 32 bit backward compatiblilty combined with some 64 bit performance.

Secondly Itanium is not waaaaayyyy of. It is actually availble to buy in HP workstations for one. Albeit for $3500.

Thirdly, I do know that DDR 333 is on the way. You have misted the point. The design of K7 is that it deal with that amount of Data being pumped into it. The whole point of DDR 333 is to increase available memory bandwidth. We have seen that the Athlon becomes saturated, ie cannot deal with anymore at 2.1 Gb/s with.

So I am sorry if you misunderstood what I was saying, maybe I wasn't clear.


Thirdly, I will readily admitt that I do feel some bias towards INTC. With every machine I have bought with an Intel chip, each one has performed flawlessly, never had any problems beyond those you would associate Windows. I did try buying a pair of AMD machines, Athlon 750's. Both were a disaster. Refusing to run stably, at one point a whole 128 MB DIMM was toasted, I lie not. They crashed on a regular basis. Whether this was AMD's fault or not is immaterial. The point is that I have bee using INTC boxes for yearsm never having any trouble. I bought Athlon and all hell was to pay.

I think I have said enough

 

[Matisaro]

Charle you are begining to look like an intel troll.

he didnt say itanium, he said mckinley, do you even know what that is? The current Itanium sucks, it is being owned by alpha processors(too bad intel bought them eh). Mckinley is supposed to be much better and it is on a .13 micron peorcess I believe. (think northwood for itanium). The current itanium runs so hot it makes a 1.5ghz tbird look like an ice cube!

The fsb of the k7 can be increased, Mine is running at 300mhz (150ddr) the k7 can be easily modified to run at 333 fsb and when ddr 333 is released you can rest assured it will be. The k7 barely uses all of its bandwidth now, it does not rely on tons of raw bandwidth to make up for its [-peep-] l1 and l2 cache like the p4.

You may feel whatever bias you want, but dont try to come off as being objective when you obviously arent. You are new to this forum and every post I have read from you so far has been an AMD bashfest. So if you do not wish to be relegated to the role of entertainment, like FUGGER, meltdoofus or juin, straighten up and fly right soldier.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

 

[charliec2uk]

He wasn't specific.

And I am sorry if that is what I sounded like. I must have gotten a little excited in the heat of the mommet. Make no mistake. I respect Advanced Microdevices. They obviously have some very talented engineers, much better than I'll ever be. (I reckon I'm doomed to be a banker just like the rest of UK engineering graduates). The K7 has proved itself a worthy rival to P3/P4. It is just that I feel that the whole Sledgehammer thing is a return to the dark times of K5 etc.

So there, I hope I have redeemed myself.

 

[Matisaro]

Its ok, just be careful how you come off and you will be respected here. Take raystonn, he is uber intel loyal, yet he always comes off as intelligent and he is respected.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

 

[somerandomguy]

Woah! You're right. I just had a look at the price of the 2Ghz P4. Ouch. Even without RDRAM, ouch.

"Ignorance is bliss, but I tend to get screwed over."

 

[FatBurger]

1. A transition of HOME USERS (which, strangely enough, is what we are), requires the ability to have good performance in 32- and 64-bit. Itanium does not provide this. Raystonn has stated (albeit vaguely), that McKinley will do this sufficiently. Nobody knows for sure, but he's rarely wrong. The Hammer line by nature will do this. We also do not know if that is the end of the line for AMD, or if they have future designs in store, or that they will decide on different designs down the road.

<font color=red>***ATTENTION***
I am NOT talking about Itanium vs. Sledgehammer, I am talking about the HOME USER, McKinley vs. Clawhammer. I know this is a ways off, but it is there. If you even THINK about using the work "Itanium" in reply to this point, I will hunt you down and cut off your testicles with a blunt, rusty saw. Thank you</font color=red>

2. AMD is being very quiet about the Hammer. Not much is known, since there are no official samples around. There are rumors of some floating around (or soon to appear), but I'll believe it when I see it. Anyhow, the reason for AMD's silence is unknown. Maybe they want the Hammer to fall suddenly and devastatingly, maybe they're embarrassed of it's performance, and maybe they just aren't far enough along to have much information for the masses yet. Who knows, but at any rate, we cannot make assumptions on it's performance.

3. A little-known factor of IA64 vs. x86-64 is the assembly code. All IA64 processors will have new assembly. You think only Microsoft's employees working on Windows will need to worry about assembly? Nope. All game engines are a significant amount (or all) assembly language. A friend of mine who is currently developing his own game engine alerted me to this fact. He is really hoping that IA64 doesn't completely obliterate x86-64, since he doesn't want to spend a year or so looking at books learning a new assembly language.



Hope this post helps out.

<font color=green>I post so you don't have to!
9/11 - RIP</font color=green>

 

[charliec2uk]

Thanks, I stand corrected.

 

[FatBurger]

What did I correct you on? I didn't realize I was going to.

<font color=green>I post so you don't have to!
9/11 - RIP</font color=green>