P-4 506 overclocking

Just bought a barely used 506 2.66 533 bus LGA775 P-4. This is the new EM64T type that supports 64-bit computing.

This is the one I got:

http://processorfinder.intel.com/sc...cFam=483&PkgType=ALL&SysBusSpd=ALL&CorSpd=ALL

It's for a co-worker buddy that is building a semi-budget computer. We plan to run it at about 3 gig or so. Well, figured I might as well see what it would do. So, I popped it into my "green" rig (see signature) and took it for a spin. Wouldn't boot at 4.2 gig, got into Windows at 4.1 gig, but was benchmark unstable, however it seemed to like 4 gig (200 FSB) ok. Not bad for an inexpensive 64-bit CPU.

wondering do those processors have HT on them?

I dont think the 533fsb chips have HT period. Nice clocks btw BB

Wish it did have HT, but xTrEmE is right, the only 533 bus CPU with HT was the old 3.06 Northwood. It's a shame.

Wow, nice performance for such a chip! Darned if these new threads aren't tempting me to upgrade from my s478 3.0C rig. -- Paul

Don't be afraid to crank the voltage up on these. Running one at 1.5v on water is pretty easy to do. Mine does like 4.4ghz or something on 1.5v so give that a spin. The only thing I don't like about these chips is until they get over 220 fsb they can't keep up with lower clocked but much higher FSB p4s. So once you get near 4.5ghz the superpi times start dropping like rocks.

i was thinking of getting this chip, 4ghz is always nice off a chip that only cost something like $140 CAD. It's also a tempting chip to buy because of its EMT64. I still dont know if i should get it or not. Batboy, have you tried anything else to get higher clocks?

o and btw, on www.infonec.com it says that the chip DOES have HT, could i be looking at a different chip? model #BX80547PE2667EN. omg if this thing has HT, im getting this chip for sure

Pretty sure Intel only enables HT on the 800 bus Prescott versions.

I didn't try too hard to go higher. This CPU was going into a buddy's system and he had already paid me cash, so I did not crank the voltage up much nor did I spend much time playing with it. I probably could of gotten at least 4.1 gig stable if I pushed more vcore and tweaked stuff more. Remember, this was on air.

thats a nice OC, 4.1ghz will definitely be satisfying for its cheap price. Are you 100% sure about the HT, everywhere i look it says that it does have HT

Every CPU I've used that has HT says "supports hyperthreading" right on the box. The 506 does not say that on the box.

On the Intel s-spec page, there is nothing listed for HT on the 506 that I tried (SL8J8).

http://processorfinder.intel.com/sc...cFam=483&PkgType=ALL&SysBusSpd=ALL&CorSpd=ALL

I picked a LGA775 3.0 CPU at random and looked at the specs. Intel says it has HT.

http://processorfinder.intel.com/sc...cFam=483&PkgType=ALL&SysBusSpd=ALL&CorSpd=ALL

hmm indeed. I think im just gonna double check this weekend at the actual store, i don't see why they'd be giving out false information. maybe just a typ-o, ill let you know as soon as i find out,

From newegg's description...

newegg said:

Hyper-Threading Support No

Click to expand...

http://www.newegg.com/Product/Product.asp?Item=N82E16819116224&ATT=Processors&CMP=OTC-streetprices

I am pretty sure batboy is right, the only HT 533fsb was the old 3.06 northwood.

batboy said:

Pretty sure Intel only enables HT on the 800 bus Prescott versions.

Click to expand...

Yep, only high end stuff gets HT.

All 800fsb CPUs, aside from the 820 & 830, and all the EEs have HT too.

P4 506 strange behavior

I've seen a lot of overclocked P4 506's here, and although I have costa rica version(that I was told is good for OC), it goes only from 2.66 to 2.8(I have read this somewhere too- that some types of this CPU go like this).
When I overclock with MSI Corecenter (my board is MSI NEO 945 with Corecell), it goes to any freq up to 187 MHz FSB (3740MHz core), but it immediately causes the CPU to brake every few seconds (even mouse pointer stops ) and oscillate FSB to double and half and again (e.g. it switches betw. 133 and 266 any second or 150 and 300...).
Although this is slightly my eleventh MB and CPU, I have never seen anything like this beefore.
Any ideas or similar behavior, boys?
Majklsson
P.S. sorry if anywhere was this discussed before...

majklsson said:

I've seen a lot of overclocked P4 506's here, and although I have costa rica version(that I was told is good for OC), it goes only from 2.66 to 2.8(I have read this somewhere too- that some types of this CPU go like this).
When I overclock with MSI Corecenter (my board is MSI NEO 945 with Corecell), it goes to any freq up to 187 MHz FSB (3740MHz core), but it immediately causes the CPU to brake every few seconds (even mouse pointer stops ) and oscillate FSB to double and half and again (e.g. it switches betw. 133 and 266 any second or 150 and 300...).
Although this is slightly my eleventh MB and CPU, I have never seen anything like this beefore.
Any ideas or similar behavior, boys?
Majklsson
P.S. sorry if anywhere was this discussed before...

Click to expand...

The motherboard has a lot to do how these processors overclock. I sold batboy the 506 he tested. It did not overclock that great with my DFI LP 875P-T, it only reached about 3.6 stable. I don't know why because the 650 that I replaced it with is running solid at 4.42G at 1.475V. I think the processor has a little more left but the vcore starts bouncing like crazy with anything higher than 1.5V.

Right, MSI mobos are usually only good for a modest O/C. That same 506 will probably go higher in a Asus or Abit mobo.

100% agreed. MSI boards, while having a fair number of supporters (due to the fact that they are cheap and glossy), are nearly complete junk. Only the absolute garbage of the world (like Amptron, ECS, Machspeed, Tekram, et. al) were worse in practice. I fell prey to the spec-sheet attractiveness of MSI's a few times, and each time withdrew nursing the financial wounds inflicted by the increased DOA rates, bugginess-during-build, failure-in-field, and increased support costs that invariably result from reliance on MSI motherboards. OC'ing makes things that much worse.

Abit boards have served me really well lately, and they are nearly as inexpensive as MSIs. I bought my current AG8-V based on the strength of the PWM power supply section, and have not regretted it. And after several AI7s, an IS7, and a BD7-II, it's clear to me it's not just a fluke. While the AG8 is not what I would recommend for you, the 945p Abit (AL8 or AL8-V) is. Quality four phase power supplies of adequate size (four FETs per phase) are necessary for the power demands of Prescott-core CPUs, especially when OC'ing enters the mix. Abit is about the cheapest board I've seen with consistantly excellent power supply sections, and they don't have many other shortcomings lately to detract from that fact.

Abit boards also boast adequate over-voltage options and the most flexible and powerful BIOSs. Abit basically invented the modern OC-friendly BIOS, and to date nobody else has taken the concept further. When you consider that Abits are also routinely the fastest board of all clock cycle per clock cycle while still being amongst the most stable, they are a fairly ideal choice for the budget-conscious buyer, enthusiast or not. The only thing I see that I like better are these new Asus 8 phase designs, but they really shoot the budget in the process. And I like essentially everything else about the Abits better, anyway.

I remember my old Abit ST6, It was good runner. One of 666 Celerons went to 1100 with box cooler on that years ago.
But my last board before this MSI was MSI 865 PE Neo2 and it seemed to be quite good board for me.
But looking on your reactions I'm going to test this 506 on other board somewhere, and I am curious.
Maybe Intel started quietly to implement protections against OC, as I read here.... http://www.analogzone.com/pwre0908.htm