is this guy correct??

[DirtyApe]

http://www.steampowered.com/forums/showthread.php?s=&postid=1406619#post1406619
thats a link to the steampowered forums(dont laugh) and this guy is claiming that a celeron D 320 can out perform a P4 3.2EE.

 

[Lazn_Work]

In what? A 4ghz celeron will have a higher MIPS than a non overclocked 3.2EE, but in any real world use it will be slower. In otherwords with a loop of small instructions that fit in the celeron's cache it will be faster than the EE, but in real world use the smaller cache of the celeron will be used up, leading to a slowdown compared to the EE that is still working out of it's larger cache.

==>Lazn

 

[Josh_B]

In what? A 4ghz celeron will have a higher MIPS than a non overclocked 3.2EE, but in any real world use it will be slower. In otherwords with a loop of small instructions that fit in the celeron's cache it will be faster than the EE, but in real world use the smaller cache of the celeron will be used up, leading to a slowdown compared to the EE that is still working out of it's larger cache.

==>Lazn

Yep, and considering the P4 is a cache-deprived, deeply pipelined CPU, having less cache hurts it badly in most situations.

Even a Duron is probably faster than all but the newest, highest clocked Celerons.

 

[ra-odin]

I think he's also forgetting you could overclock the P4 as well and beat the pants off that Celeron

 

[Dan_D]

Yeah a Celeron D has a Prescott core, just like the P4A's. I'll tell you why all these chips are crap.

They have all the Prescott's weaknesses and none of it's strengths. They've got the heat, the slow clock for clock performance, and a 533MHz bus. On top of that neither has Hyperthreading.

The Pentium 4 Prescott A's have the same L2 cache but that's not enough to get past the rest.

The Pentium 4 Prescott "E"s have the 800MHz FSB, the Hyperthreading (somewhat improved on over the "C" due to the longer pipeline) and since they are offered in speeds of 3.4GHz plus they finally can start to overtake the Northwood's performance.

Newer models even have the NX bit and EMT64. LGA775 only of course.

The 3.2EE is also a Northwood. So less heat, faster clock for clock performance, 512K L2, 2MB of L3 cache and it's a good overclocker. I'd bet that even with the Celeron D at 4GHz, the 3.2EE would still beat it most of the time.

 

[EnderW]

Whoa whoa whoa, my bad.

I just called my bud to confirm his CPU, and he does have the 3.2E, not the EE like I thought it was.

MY BAD, TOTALLY MY BAD.

 

[Stugots]

that guy said his buddy with the 3.2ee with the same mobo and ram as he was getting under 18000 in 3dmarks01, and his system with the celeron D @ 3.85 was getting 18922. my system pulls around 20000k with my cpu @ 3.4ghz. it could be faster in synthetic benchmarks and maybe 3dmarks but id much rather have the 3.2ee over the celeron D.

but on the other hand, a system with a 3.2ee barely hitting 18000 doesnt sound right.

 

[dnas]

Yeah a Celeron D has a Prescott core, just like the P4A's. I'll tell you why all these chips are crap.

They have all the Prescott's weaknesses and none of it's strengths. They've got the heat, the slow clock for clock performance, and a 533MHz bus. On top of that neither has Hyperthreading.

Not true regarding heat........

The fact that is has only 25% of the L2 means that Celeron Ds are not hot chips. I've run a Celeron D 320 (2.4G) at 3.6G, 200MHz fsb, although not stable. It runs perfectly at 3.4G, stock fan and voltage, 41C idle 52C under load (Prime95) I would venture to say that it would beat the 3.2EE stock speed in MPEG2 encoding, however that would be about the only thing it would do faster.(look at xbitlabs.com for their review on the Celeron D)

So it is not crap.

 

[operaman]

It is a great value chip for sure, but a P4 EE will beat it in most balanced benchmarks...