celeron vs p4 wtih pc133 sdram

[omeryounos]

How does a celeron 1ghz performs against pentium 4 1.6 ghz with pc133 ram,or pentium 3 1 ghz in games.

------------------
intel Celeron 1GHZ
512 mb sdram
asus v7100 pro
(geforce 2 mx 400)
40 Gb Ultra ata100 Hd
17'' Monitor.
Windows XP Pro

 

[dhlucke]

Which P4? Overclocked chips?

<font color=red>God</font color=red> <font color=blue>Bless</font color=blue> <font color=red>America!</font color=red>

 

[omeryounos]

not overclocked .stock p4 1.6 ghz socket 478 with intel 845 mainboard

 

[NurseMSIC]

Well mine IS o/c, so cant help you exactly. But i see that we have same (ish) graphics card, so get yourself some benchmarks and go here::

For PC Mark
http://service.madonion.com/compare?pcm=222630

For 3D Mark 2001 SE
http://service.madonion.com/compare?2k1=3278868

and just for reference, here was my P3, a 733 with PC133::
http://service.madonion.com/compare?2k1=2558948

The spirit of Monkey was ...... irrepressible

 

[omeryounos]

hey doesn't any one else know this question .y no answers.

 

[omeryounos]

hey no answer . im surprised. where r the gurus????

 

[buddry]

None of us gurus have celerons

If you can find it on the internet, it HAS to be true!

 

[ath0mps0]

And no respectable guru is going to have a P4 with the i845(SDR) chipset and PC133 RAM. That's like castrating your Thoroughbred Stallion... :tongue:

If the thought I thought I thought had been the thought I thought, I wouldn't have thought so much.

 

[FatBurger]

Give it some time, n00b.

But this is a pointless question, you might as well ask whether a Geo Metro or a Chevy Sprint is faster.

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft

 

[Crashman]

OK, first of all, the Celeron 1GHz is about as powerfull as the PIII 750. The PIII 1000EB is about 33% better than the PIII 750, and about the same performance as a P4 1.4 with RDRAM. A P4 with SDRAM is about 30% max slower than a P4 with RDRAM. So...A P4 1.6 with SDRAM is about the same as a P4 1.2 with RDRAM would have been (if it existed), which would make it about the speed of a PIII 866, which is much faster than a PIII 750 (up to 25% because of it's faster memory)...I'll guess the P4 1.6 with PC133 is about 20-25% better than a Celeron 1GHz. Geeze, that really makes the P4 look bad. But hey, we knew that SDR SDRAM on the P4 sucked.

As for the PIII 1000EB, well, it beets the 1.2GHz Tualatin Celeron in Games when tested on my system. In fact, out of all those combinations, the PIII 1000EB is the fastest in Games. Unless you're using games that the P4 was optimised for (Quake 3 and probably some other obscure Open GL games).

What's the frequency, Kenneth?

 

[FatBurger]

My brain hurts

<font color=blue>If you don't buy Windows, then the terrorists have already won!</font color=blue> - Microsoft

 

[eden]

Oh brother, not another Quake 3 optimized for P4 claim....

--
Thunderbirds in wintertime, Northwoods in summertime! :lol:

 

[Crashman]

LOL...NO! More like the P4 optimized for OpenGL. It seems kind of funny that no matter how poorly built a P4 system is, it still does OK in OpenGL.

What's the frequency, Kenneth?

 

[Crashman]

I guess everyone was too lazy to combine previous finding as exposed by THG with personall experience and simple mathmatical estimations.

What's the frequency, Kenneth?

 

[eden]

Well it does, but if people see 320 frames done and say it's optimized, god knows why they still study CPUs. If you look clock per clock, the Athlon wins, which means it is not optimized to be better than the competition. It may be optimized because without that it is actually over 30% below per clock, and the optimization brings it 5% less or so, but please don't go assume it is optimized to run better than Athlons, when it's clearly and only a clock speed advantage, same as Internet Content Creation tests being raw clock speed dependant.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol:

 

[Crashman]

Did I say anything about Athlons?

What's the frequency, Kenneth?

 

[eden]

Your words describe your spiritual feelings, lol.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol:

 

[Crashman]

? OH, you mean when I said the PIII 1000EB would be faster than a P4 (assuming Willy) 1.6 with SDRAM? Well, most people knew the P4 sucked, the Northwood was only a slight improvement, and the only reason a P4 can beet any Athlon is because AMD refuses to release faster cores in a timely manner. It would be handy if the Thoroughbred were actually available to see how far it goes, but then he didn't ask about AMD, he didn't even ask about a P4 with RDRAM. He ask about the worst performing clock for clock P4/SDRAM combination against the second worst performing clock for clock Celeron against a moderately well performing but clock limited PIII. Perhaps he's looking at some very low end systems from an Intel Only builder, probably an OEM. Or perhaps he was looking at the comparison from an academic standpoint. At any rate, an inexpesive XP system, even paired with SDRAM, would obviously outperform any of those solutions, at a similar cost to the P4/SDRAM system. As for me, I'm building an overclocker, I'm using the right parts.

What's the frequency, Kenneth?

 

[eden]

Oh no I wasn't hinting at that. It's just the claim for P4 Optimized for Q3 is sometimes very overused.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol:

 

[Crashman]

How about this "the original P4 was around 30% behind the Thunderbird, clock for clock, but Open GL optomization narrowed the gap to less than 10% in at least one game, Quake3, and possibly other OpenGL apps".

What's the frequency, Kenneth?

 

[FatBurger]

If you look clock per clock, the Athlon wins, which means it is not optimized to be better than the competition.

That's not true at all. If it was, then every program on the market is optimized for the G4, since it does the most per clock.

<font color=blue>Hi mom!</font color=blue>

 

[eden]

<A HREF="http://www6.tomshardware.com/cpu/02q2/020402/p4_2400-08.html" target="_new">http://www6.tomshardware.com/cpu/02q2/020402/p4_2400-08.html</A> (the closest match to a Northwood I could find, is the 1.6A, look at the XP2000 1.66GHZ against it, roughly 30 more frames even if it is just 66MHZ above. Give or take 10 frames, it's still better.
I couldn't agree more with you on the G4, but again it all comes down not just to IPC, but to its instruction optimizations and such. Look I had described once before about a corn field example and how an optimized one would be the one that really does more per field acre. If you want me to dig my LONG example up, I will.

My main point is that yes the P4 MAY be optimized, but if it isn't more per clock, (again this is one of the only areas I can possibly talk about with clock per clock refering, otherwise overall performance is no longer this matter), then why is it optimized? We're discussing PC btw not Mac. If in the x86 home user world, an Athlon does better per clock even without that special Q3 patch, then again it is better than P4, but not even optimized. If the P4 was above per clock because of the optimizations, I'd say it is "Heavily optimized" to "optimally" run on the P4 more than Athlon. Sadly it isn't the issue, you can say it has optimizations on P4, but not that it will run optimally more than Athlon. THAT is what I want to say. However I will agree to say and admit thoroughly, Quake 3 does run better on P4 if you want to decide which processor(again, big clock speed advantage, as Q3 benefits from both card and CPU). Of course if you really want 300 FPS that is.

BTW Crashman, is it wrong to change subjects and debate? This has happened tons of times here, so why should I always be the one not allowed to change topics with someone who said something I'd like to debate? Geez lighten and loosen up guys, I was just trying to prove something and have some fun debating!

EDIT: Ok I just found another good example. You think Quake 3 is optimized better for P4, you go buy a 1.5GHZ P4 or 1.6A if you want, then you find out a similar clocked AXP, does better. So much for the optimized P4 huh?
--
Thunderbirds in wintertime, Northwoods in summertime! :lol: <P ID="edit"><FONT SIZE=-1><EM>Edited by Eden on 05/01/02 09:47 PM.</EM></FONT></P>

 

[Oni]

Crashman what do you plan to put into your P4 overclocking rig?

Gosh I'm such a nerd sometimes, but then again arn't we all.

 

[FatBurger]

I still don't get why you think work done per clock means something is better optimized. The Itanium could possibly get more done per clock in x86 code than the P4, so that means that programs are more heavily optimized for Itanium? Not in the least.

Try this. Grab Sandra numbers (heavily synthetic) for P4 and Athlon, and then look at a single program's benchmarks for those. Compare the % difference between the synthetic and real-world benchmark, and that'll tell you how optimized it is. That probably isn't clear, so let me give you an example.

Say you have a P2 2.4 and an XP 2100+. Say Sandra rates the processors as 10 and 9, repsectively (I'm pulling numbers out of my ass just to get the example across, not because I think this is what would actually be the case. The numbers themselves don't matter, just what they mean). So the XP is 10% slower (less powerful, if you prefer that term) overall. Then you run Quake 3 (or Photoshop, or Lightwave, or Divx, or MP3, or UT, or...), and they get 6,978 FPS and 4,178 FSP. That's a difference of 41% (I think, I haven't had my second cup of coffee). 10% <==> 41% is a substantial difference, so you could conclude that Q3 is more optimized for the P4.

Does that make sense? Optimizations are about making a processor <i>more</i> effective, not the <i>most</i> effective, if that helps any.

<font color=blue>Hi mom!</font color=blue>

 

[eden]

I understand you, and I did not say optimizations won't make the processor better. What I am hinting at, is even though it has optimizations, the end-user who buy a similar clocked Athlon, versus a P4 with optimizations in Q3, still gets Q3 FPS better on Athlon. This means that even if it is optimized, it still has not done better to break in the home user's experience. This is valid for Geforce 3 technology. Look at the GF2 Ultra vs a Ti200 in Aquanox. Ultra is much much higher clocked, same DDR memory. Yet even if the Xbar mem controller helps, the Ti200 manages to go high forward. Why? Solely thanks to the GF3 based opt. Now say you optimized the Ultra with DX8.1 or GF3 features, it still does not fully pass Ti200's FPS, solely and again because it is backed up by also the Xbar mem controller. That is how I see P4 in Q3 or any optimization for that matter. If a guy buys a clock per clock CPU from either AMD or Intel, FOR HOME USE AS IN X86 PC, NO IA64, NO MAC. If Q3 is optimized for P4, but even at the same clock it has not fully bypassed Athlon, I would not say it is more beneficial to get P4 anymore.

Conclusion and again: Optimization helps, for sure, but you cannot buy something that is optimized, and yet get lower performance thana similar clocked CPU! It makes no sense until those optimizations completly eradicate the other CPU's clock-per-clock performance without any opt. This would also mean that if Athlons received Q3 opt, they would just kick P4's ass and its optimizations. From there, the only way and that we all want, is to put in the FPU, increase ALU performance, cache, anything that really increases the integrity and performance. Get what I am saying?


PS: I reread your last quote, in the end you say "Not the most effective", which is what I mean. Don't go fool a person to buy a 1.6A and say it has Q3 opt, when he can buy an XP1.6GHZ (if it existed, or else a 1600+ or 1.6 Tbird), and get better performance.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: