• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Will the P4 be "outclassing" the Athlon, half a year from now?

That's a fascinating read on software optimization. I don't think the majority of programs will be written for SSE2 until early 02' though. It's unfortunate that the generic recompilation is so poor for non SSE2 when using Intel's SSE2 compiler. One wonders if it was just poor engineering on Intel's part or intentional de-optimization for competitors...cough..cough..AMD.
 
< Thanks to this new compiler and higher clockspeeds, the fastest Pentium 4 should be able to decrease or even close the performance gap with the fastest Athlon in Pentium 4 optimized workstation applications. The Pentium 4 might even increase its performance advantage over the Athlon in some streaming applications (video encoding). >

That pretty much states it...and it doesn't state much...

the increase in the sse-2 optimization helps it to overcome its raw fpu power weakness but not enough to make clock to clock comparison a reality. A significantly clocked slower palomino (the fastest of the amd chips of that category, mp in the test) will still hold its own.

I am a little disappointed I thought the optimization was supposed to possibly make a 1.5 p4 act like a 1.5 tbird, not just close the gap....

If athlon holds this, it will be soon enough that their future chip (and I can't remember the code name) will have sse-2 optimization as well and will benefit from any optimization also...
 
All I say is that this competition is GREAT for the computer market. With Athlon and Intel going neck-and-neck with each other, you'll see better and better products coming out for better and better prices. I mean, you can already see it happening. I was able to get a 1 GHz Pentium III CPU a few weeks ago to upgrade my computer, and it only cost me 240. Three years ago, a fast CPU like that would have cost me twice as much. I still say that Intel will come out on top, however. They've been in this longer than AMD, and once they get their heads out of their snooty arses, they'll realize what they need to do and produce an insanely powerful chip.
 
Check out Aces Hardware for a comparison of Microsoft's upcoming new compiler and Intel's. You'll notice that the Athlon makes significant gains with Microsoft's, while the P4 makes significant gains with Intel's.
 


<< In six months, this will be a moot point. >>


Could you be a little less clear please?😉

The Thoroughbred will have full SSE 2 optimizations. Unless Intel can really turn things around with Northwood AMD will continue to lead in performance.

Edited for a typo
 
jdurg,

three years ago, there were no GHz computers available for us low life consumers. But look at the market now, if you have below a GHz, you are falling behind 🙂
 
InsaneMorphius
But look at the market now, if you have below a GHz, you are falling behind

Damn! you wanta donate some hardware to poor ole me to replace my old slow PIII @ 820 ?😉😛
 
Wingznut Pez wrote:

&quot;In six months, this will be a moot point.&quot;

Not necessarily. If AMD does include SSE2 optimizations in a future core (as I expect they will) and Intel doesn't seriously refine the P4 core (besides just shrinking the die), AMD will blaze well ahead. But since we don't really know, it's pointless to speculate. I don't think it'll be &quot;moot&quot; in six months, however 😀
 
I think Wingznut is hinting at upcoming technologies that Intel has not announced. He's been quietly doing so for a few weeks now.😎 Of course he has been very sly about it, so it could be anything from a 1MB cache on the P4 to an on-die Flux capacitor.

😉
 
I hope its not a flux capacitor... I dont wanna have to install a motor spinning at 82 mph to get that performance boost 😉

Well, I dont think &quot;half a year from now&quot; is long enough. I'm assuming this is based on new software, and not a hardware revision. Based on historical trends from MMX, 3dNow!, etc., I think it will be at least 2 years before most apps come optimized. That is too long to wait, imho. Presently with a tbird or MP I can get performance now that will exceed that of P4's for at least two years in most apps.
 
It's also been shown that SSE-2 doesn't necessarily push P4 over the Athlon on any given application. That's why I made the comment that, unless Intel makes significant changes to the P4 core (other than going to .13u), when AMD offers a core with SSE-2 there will be no comparison.
 
I meant that SSE or not, the new P4 will rock.

(My apologies for being vague... It made sense to me when I wrote it. 😉 )
 
When it comes to cpu power we are spoiled. 1.4GHz, 1.7GHz, P4, Thunderbird, Palomino and so on. I have an AXIA and yes it is fast. But let's put things in perspective. I just built a Duron 750@1000. The thing only costs $38 and runs Windows 98 fast. It plays games just as well as my AXIA, for the present it is all we really need!!! It runs really complex programs just a tad slower, but not much. As cheap as RAM is I threw 256MB of PC-133 in there and it multi-tasks quite well. I almost feel guilty using anything more, can we really justify the extra money other than to feed our own selfish needs?
 
1 thing though ,the T birds are way faster than the Durons at SETI ............................. what do you mean that's not the most important thing!?:Q..................😉
 
Wingznut PEZ: You work as a fab guy for Intel, right? So how does a lowly guy like you that wears a static suit all day know about all these future techs? Not to insult you or anything, I'm just wondering how you learn these things. Or do they just trickle down from management to you guys?
 
I hope its not a flux capacitor... I dont wanna have to install a motor spinning at 82

That's 88 mph, and don't forget the 1.21 gigawatt power draw. Definitely going to need an Enermax . . .

And, correct me if I'm wrong, but Wingnut isn't a bunny suit guy; he's working on the Itanium team, Swanny. So he probably does know of what he speaks . . .
 
Xerox, I think you have me confused with pm. He's a circuit designer with the McKinley team. I'm a Lithography Technician on the Tualatin and Northwood processes. Also with R&amp;D beyond those.

Swanny, I learn these things in many, many different ways.

The #1 way that I learn things is that I work very closely with the process engineers. They get an idea, I implement it, and then we go over how well (or not well) it worked.

I also attend meetings, in which updates are given on how the processes are performing, as well as what's coming &quot;next&quot;.

And we also get updates almost daily (via intranet and/or email) of how the new processes are coming. (Die yeilds, bin splits, RISO targets, etc...)

Sure, there's a lot of things that I'm not told. A good example would be the Tualatin needing a new motherboard. That's something that's irrelevant to my job, so it's not necessarily shared with me. The &quot;need to know&quot; policy is pretty strong at Intel, basically to help keep private matters private. Heh, most people assume that I know more than I really do about future technologies. pm and I laugh about that.

And I do think that I know more about these things than other Techs. But that's because I'm interested and I try to seek out the answers whenever possible. But that's no different than any other job.
 
Wingznut PEZ: Ok, I guess from other threads I though you were lower in the &quot;food chain&quot; than that. Now I'll never doubt you agian😉 Not that that I would anyway😎
 
WingznutPEZ
Tualatin needing a new motherboard.

Is that the new PIII on 0.13 micron process? ,if so I thought some older boards (including BX boards) could run them.Abit has bios updates for the BX 133 boards which include support for thr Tualatin.😕
 


<< Is that the new PIII on 0.13 micron process? >>



Yup, Tualatin and Coppermine-T will both be running on a .13 micron process (I think). You may be right about some older boards supporting the Tualatin, but all of them should support the Coppermine-T.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top