Why Do Pentium 4 processors overclock better than Athlon XPs?

[magomago]

I'm just wondering this...Everyone got excited (including me) when Chiz pushed his 1.7 T-bred B to 2.4Ghz (that is VERY cool ) but you got 2 Ghz P4s breaking the 3Ghz Barriers...you got 1.6s hitting 2.4-6

Is the P4 architecture more "friendly" to overclocking? Just wondering

 

[OverVolt]

I think the P4's have more headroom since they run cooler. Also P4's .13micron has gone as high as 4Ghz where AMD's same process is struggling to reach past the 2.3Ghz mark. Only thing is... P4's can oxidize or somethin under voltages higher than 1.8v ish while AMD has to worry about heat buildup.

I think AMD is desperate to gain performance by adding the 512k cache to the Tbred-B "Barton" which only helps the CPU about 5% on average. The extra 17million transistors will really hamper overclocking on Barton since the cpu can only run as fast as the slowest transistor (I think correct me if i'm wrong on any of this).

 

[formulav8]

Originally posted by: OverVolt
I think the P4's have more headroom since they run cooler. Also P4's .13micron has gone as high as 4Ghz where AMD's same process is struggling to reach past the 2.3Ghz mark. Only thing is... P4's can oxidize or somethin under voltages higher than 1.8v ish while AMD has to worry about heat buildup.

Umm, I think you are WAY off. The main reason I can see is that the P4 has twice the amount of pipeline stages, 10 vs 20. That is one of the main reason the Athlons IPC kills the P4's.


Jason

 

[BFG10K]

Why Do Pentium 4 processors overclock better than Athlon XPs?

The same reason they have a higher frequency to begin with. Their design centres around attaining a high clock speed and hence they have more headroom when you try to go above their stock speed.

 

[OverVolt]

Sry! I don't really understand the long vs short pipeline thing between the two, all i know is P4 has the long pipeline + massive frequency + lower IPC while AMD has short pipeline + lower frequency + higher IPC.

I don't see how they affect it's overclockabilty- oh well!

I always thought to get a good overclock you needed good cooling and extra voltage for stabilty plus watching your pci/agp frequency's along with mem timings/voltage. but my overclocked rig just totally hosed windows with corrupted files i think i pushed it too far on a crap mobo.

 

[DannyBoy]

Originally posted by: BFG10K

Why Do Pentium 4 processors overclock better than Athlon XPs?

The same reason they have a higher frequency to begin with. Their design centres around attaining a high clock speed and hence they have more headroom when you try to go above their stock speed.

I will back this up. Intel aims at the "How much higher can we make the clock speed" whilst AMD are more concerned about "How much Extra Performance can we give this thing". (Hence the XP for Xtra-Performance)

XP's IMO run at a more realistic clock speed offering the same performance as higher clocked P4's. I personally think Intel have some stupid idea that they wanna beat Moores Law :|

EDIT: Oh ya and higher clock speeds sound more attractive to the 'not so computer literate' average consumer!


Dan

 

[Cerb]

In the chip, there is a pipeline to get things done...all the stuff it needs to do go through this pipeline (it is called a pieline for a reason).
The pipeline has stages, and each one does a small amount of work.
IIRC, each stage has a single clock cycle to do its work, then it pushes it to the next stage.
So, the P4 has a longer pipeline. So that means each one can do less work, so to keep up with the power of chips with smaller pipelines (like their own PIIIs), it must reach a significantly higher clock speed, since they are doing less each cycle.
The P4's higher clock speed and wider memory bus lets it shove around more information, hence its dominance in video benchmarks; though the price is that it loses processing power to do that. There are more disadvantages to longer pipelines than just less power per clock. IIRC, if the chip has to wait for memory to give it something and the pipeline needs to be filled again, since it does less, it takes longer. This is part of why the prediction part of the chip was such a big deal going to the northwoods (something about having such high clockspeeds and not being able to get signals across the CPU in a single cycle).
The current Athlons are more on the side of raw calculation, and likely that will continue with the Hammer, since, for one thing, it gives it a niche.

here's a good link I found: Google is God's search engine.

 

[Wingznut]

Originally posted by: MrFiTTy
Intel aims at the "How much higher can we make the clock speed" whilst AMD are more concerned about "How much Extra Performance can we give this thing". (Hence the XP for Xtra-Performance)

Actually, both manufacturers are concerned with getting the highest performance. They just chose to take different routes in getting there.

As for the "XP" nomenclature, it's just a marketing effort in an attempt to align it with Windows XP. Here is AMD's response in a FAQ:

Q: What?s the meaning of the ?XP? in the AMD Athlon XP processor name?

A: The ?XP? modifier is designed to convey the extreme performance that AMD Athlon XP processors deliver for the Microsoft® Windows® XP operating system.

 

[OverVolt]

Originally posted by: Cerb
In the chip, there is a pipeline to get things done...all the stuff it needs to do go through this pipeline (it is called a pieline for a reason).
The pipeline has stages, and each one does a small amount of work.
IIRC, each stage has a single clock cycle to do its work, then it pushes it to the next stage.
So, the P4 has a longer pipeline. So that means each one can do less work, so to keep up with the power of chips with smaller pipelines (like their own PIIIs), it must reach a significantly higher clock speed, since they are doing less each cycle.
The P4's higher clock speed and wider memory bus lets it shove around more information, hence its dominance in video benchmarks; though the price is that it loses processing power to do that. There are more disadvantages to longer pipelines than just less power per clock. IIRC, if the chip has to wait for memory to give it something and the pipeline needs to be filled again, since it does less, it takes longer. This is part of why the prediction part of the chip was such a big deal going to the northwoods (something about having such high clockspeeds and not being able to get signals across the CPU in a single cycle).
The current Athlons are more on the side of raw calculation, and likely that will continue with the Hammer, since, for one thing, it gives it a niche.

here's a good link I found: Google is God's search engine.

Makes perfectly good sense to me. You could be a teacher

 

[dexvx]

Originally posted by: MrFiTTy

Originally posted by: BFG10K

Why Do Pentium 4 processors overclock better than Athlon XPs?

The same reason they have a higher frequency to begin with. Their design centres around attaining a high clock speed and hence they have more headroom when you try to go above their stock speed.

I will back this up. Intel aims at the "How much higher can we make the clock speed" whilst AMD are more concerned about "How much Extra Performance can we give this thing". (Hence the XP for Xtra-Performance)

XP's IMO run at a more realistic clock speed offering the same performance as higher clocked P4's. I personally think Intel have some stupid idea that they wanna beat Moores Law :|

EDIT: Oh ya and higher clock speeds sound more attractive to the 'not so computer literate' average consumer!


Dan

Ah yes, more myths about Moore's Law. It has nothing to do with clockspeeds, only transistor count per unit area. Its just 2 ways of solving the problem, go for super high clock speeds with low IPC, or low clock speed with high IPC? FYI, AMD doesnt have the highest IPC by far. Itanium II is the currect FP/clock speed champ.

 

[Nemesis77]

Originally posted by: OverVolt
I think the P4's have more headroom since they run cooler.

P4 doesnt really run cooler than Athlon XP does

 

[bgeh]

Originally posted by: magomago
I'm just wondering this...Everyone got excited (including me) when Chiz pushed his 1.7 T-bred B to 2.4Ghz (that is VERY cool ) but you got 2 Ghz P4s breaking the 3Ghz Barriers...you got 1.6s hitting 2.4-6

Is the P4 architecture more "friendly" to overclocking? Just wondering

it's true if you count Mhz to Mhz
but if you count them using AMD's PR rating, the numbers sure looks better(and performance is also around the same)

 

[ElFenix]

and here i thought it had to do with how well intel's .13u fabs were working, as the .18u p4s weren't all that great overclockers

 

[imgod2u]

As many have said, the design of the P4 was aimed at achieving higher clockrates. At .13 micron with 6-layers of copper interconnects, the P4's design seems to reach its maximum at 3-4 GHz vs 2.5ish for the K7 design on the .13 micron 9-layer copper interconnect fab. The main reason, as others pointed out, is the significantly longer hyper-pipelined design. See 1 and 2 for a description of hyperpipelining (you'll need Macromedia Flash Player). Also see this article for an idea behind basic design differences (it uses the P4 and PPC G4 as examples).

 

[zayened]

whats the chances the barton is going to run cooler than the tbred, as the p4 ran much cooler than the p3?

 

[BD231]

Seeing a processor do a 30 to 50% clock increase on average means the processor was ment to run at that speed and probably higher, its become a trend for me as an overclocker. I've found that more current processors are generally able to reach, or get somewhere close to their more expensive big brothers max clock. Don't look at it like the P4 can overclock better, the design of the core itself just has a different MHZ ceiling than the Athlon dose.

 

[BD231]

Originally posted by: zayened
whats the chances the barton is going to run cooler than the tbred, as the p4 ran much cooler than the p3?

I'd imagine that extra cache would bring along some added heat in itself and the core is still pretty small, if you ask me I don't think it's going to run much cooler.

 

[giocopiano]

Just think of it like different boxing weight classes.
Some of those 11 stone fellas are fast but they would need a helluva lot of punches to down a heavyweight. If they survive that long.

Anyway, Thunderbird B still overclocks enough to the point where its performance can outpunch similarly rated P4s depending on application.

 

[alexruiz]

Originally posted by: zayened
whats the chances the barton is going to run cooler than the tbred, as the p4 ran much cooler than the p3?

I don't think the P4 runs cooler than the P3.... I know for sure than in fact, the P4 is the most power hungry CPU (3.06 GHz eats over 100 W of ELECTRICAL power)..... It is more efficient disipating the heat than the Athlon, and that is why it is said to run "cooler", but no way it can run cooler than a P3..... or have you seen ultra slim notebooks or rack mount servers using the same CPU density for a P4 than a P3.

Can anyone confirm this???

 

[paralazarguer]

The P4 does run much cooler than the Athlon XP. People will say that they both run as hot but they don't. Take an Athlon XP that dissipates 80Watts. That's how much power is being dissipated for the whole chip.
Right.
So, let's say that you also have a P4 which dissipates 80Watts. That's also how much power is being dissipated for the whole chip.
So...
If both chips had the same die size (ignoring heat spreader) then they would run equally hot not counting any architectural differences etc. But they're not the same size. The die size of a P4 is much larger than an Athlon's die size.
It dissipates more heat more quickly.
Just like anything, the more surface area you have, the faster it will dissipate heat.
So, while a given P4 and XP might each run at 80Watts, the P4 will run decidedly cooler give no heatspreader and the same cooler.
It's physics.

 

[ElFenix]

Originally posted by: paralazarguer
The P4 does run much cooler than the Athlon XP. People will say that they both run as hot but they don't. Take an Athlon XP that dissipates 80Watts. That's how much power is being dissipated for the whole chip.
Right.
So, let's say that you also have a P4 which dissipates 80Watts. That's also how much power is being dissipated for the whole chip.
So...
If both chips had the same die size (ignoring heat spreader) then they would run equally hot not counting any architectural differences etc. But they're not the same size. The die size of a P4 is much larger than an Athlon's die size.
It dissipates more heat more quickly.
Just like anything, the more surface area you have, the faster it will dissipate heat.
So, while a given P4 and XP might each run at 80Watts, the P4 will run decidedly cooler give no heatspreader and the same cooler.
It's physics.

no, the 80 watts is the amount of heat being dissappated, they both run just as hot.

 

[cmdrdredd]

I think that perhaps Intel boards (which is what everyone uses for overclocking) are more stable at high FSB than most AMD boards (VIA and SIS). However, I've seen many people go real far with their Nforce2 platform.

 

[chizow]

Originally posted by: ElFenix

Originally posted by: paralazarguer
The P4 does run much cooler than the Athlon XP. People will say that they both run as hot but they don't. Take an Athlon XP that dissipates 80Watts. That's how much power is being dissipated for the whole chip.
Right.
So, let's say that you also have a P4 which dissipates 80Watts. That's also how much power is being dissipated for the whole chip.
So...
If both chips had the same die size (ignoring heat spreader) then they would run equally hot not counting any architectural differences etc. But they're not the same size. The die size of a P4 is much larger than an Athlon's die size.
It dissipates more heat more quickly.
Just like anything, the more surface area you have, the faster it will dissipate heat.
So, while a given P4 and XP might each run at 80Watts, the P4 will run decidedly cooler give no heatspreader and the same cooler.
It's physics.

no, the 80 watts is the amount of heat being dissappated, they both run just as hot.

Not to mention heat isn't uniformly distributed on the core. Both AMD and Intel have said numerous times that most of the heat on the core are generated in a few areas, a fraction of the total die size.

Anyways, I always felt it was more an overhead/stepping issue. Naturally, lower "rated" CPUs will have more headroom if they share a similar stepping/batch as a higher rated CPU. Isn't that why we run around and chase revisions and steppings? If you are talking about how that translates into actual MHz between a P4 and XP, I'd say it would be a function of the P4's ability to scale to high clock speeds as well as Intel's superior fab capabilities.

Btw, I'm running at 2380MHz now, as 2430MHz was a bit too hot (and too loud) for only a 50MHz increase. So I guess its not as cool anymore I'm still pretty excited about it though! I'm looking at the 2500+ Barton; I wanna see what those things can do

Chiz

 

[Cerb]

I think that perhaps Intel boards (which is what everyone uses for overclocking) are more stable at high FSB than most AMD boards (VIA and SIS). However, I've seen many people go real far with their Nforce2 platform.

On the VIAs, yeah, they're picky
The SiS ones are fine, like the K7S5A, as long as you redo the chipset heatsink. Use some AS adhesive and a good heatsink and people reach those high FSBs just fine.
On the other hand, the NF2 does have infinitely better sound, the NIC uses less CPU, etc., an you don't have the mod any parts of the board to OC the bus. Oh, and dual-channel DDR...

 

[grim122]

different core design, architecture that because of its design has a lot more headroom than the xp, and a really big multiplier (small increases in FSB x a big multi makes for larger jumps in freqency than the xp