3.2c SL792

[microfire]

Can't find much info on this cpu in various forums, its not even on intel's sspec finder.
Underneth the chip it had a even pile of resistors like the EE chips.

Not sure where its going get to just yet, as much as possible I hope. :p Only got the chip it today. Haven't had time to do any real testing just yet.
Looking ok at stock vcore at 3.6ghz.
3.7+ needed a slight bump in vcore to keep it happy when playing games.
3.8 is not looking so good maybe more voltage could solve this. 3.9 would crash if I tryed to play a game or just reboot at any moment on the desktop.
Most I tryed was 4ghz @ 1.675v, I could use the bios but it only got half through a windows boot.
Im using a Zalman 7000cu and ASC paste. Only put a matchhead size paste on there, was very thin spread out, ive got to question if I used enough(I always do, if not the other way around).

Memory bandwidth not looking good, I was hoping to break 6000mb/s at the higher fsb I tested but the best I could do was 5700mb/s while playing with abits overclock utility.

At 3.6ghz 225mhz 800strap 2-2-2-5 sr-a-a-d-d 1:1 it only pulls 5600mb/s :(

Came with a weird heatsink that fins coming out from the center north, south, east and west, with a copper dot right in the middle. It actually looked like a decent heatsink. Wonder if it better than my old TT spark 7 that I dont use now.

Anyone else got this cpu and how have you been going with overclocking it?

[TC]

I just got this chip from Newegg as well and started hunting since I also noticed it's not in Intel's spec list. I wanted to see if it's an m0, and I'm assuming it is, but does anybody else have an idea how to tell? BTW I got it to replace my 2.8E prescott, and I'm planning to put it into my vapochill tonight if I can get finished with work before too late.

[microfire]

yea its hard to tell, visually it looks like a m0 based on whats underneth the chip. With cpuZ it says D1 though.
I was about to buy prescott, thats what I was wanting. Heat didn't bother me much until it looked as if the stress might be to much for the motherboard.

[TC]

Yep the base has the same full set of caps that I've seen pictured on the other m0's. Really odd that Intel doesn't have it listed since they already have the prescotts. Mine had a pack date of 1/27 - what about yours? I'm hoping I can take this up to 4.1-4.3 in my vapo.

[microfire]

FPO/BATCH # : L336B456

PACK DATE: 12/03/03


Im having some big problems pass 230fsb. Not sure if the cpu can give me memory errors with memtest86? but im getting tons of them in test 5. No matter if I change the timings, change to strap around 5:4 3:2, or even if I switch around the modules or use one stick, ive been using 3.2v in all tests.
I tryed as high as1.65v plus 4ghz / 250fsb 5:4 but memest starts going crazy on test 1.
I know 4ghz is aiming to high on air but errors at 230fsb and over 5:4 is just insane with errors. If ity really was the cpu then I dont think it would post.
Its not looking good at all if im going to running at 3.5ghz for stable operation is just a major disappointment if it is the fault of the cpu.

[TC]

Yeah I think you can get memtest errors from the cpu, but don't hold me to that. Wow that's kind of disappointing if you really are hitting a wall at 230. I decided to buy the 3.2 instead of 3.4 thinking there wouldn't be a huge difference, but I'll dump this one if it turns out to be a loser. Maybe it's so bad Intel doesn't want to acknowledge it exists :D

[TC]

Slapped it in the vapo and I'm not real impressed. It posts around 4100, but won't load windows until I drop it down to 4 at 1.65 volts.

[microfire]

Your right, I never thought there would be a huge difference between the 3.4 and 3.2, the biggest difference for me was the price, also I was hoping on fully tapping out my memory with the 16 multiplier.
All is not lost though, ive now got the voltage to 1.675 which give me 1.65v actual, considering default voltage on this chip is 1.55v I havent increased the overall voltage much more than stock. This chip responds far differently than my 2.4b C1 chip in the way it overclocks. I dont think im going to be affraid to throw even slighty more voltage at it if I have to.
Right now got it at 3.75ghz / 235fsb with the above voltage, no memtest errors. If I drop the voltage down the errors will return, so yes the cpu can give memtest errors by the looks of things. I don't think that I will get much more than this, but if I could get 3.8ghz on air I would very happy at that, im close enough already.
Maybe you got a dud chip, more dud than mine, im not to sure about mine I still have to test it more. Have you tryed yours on air to see what it will do. If your using a vapo wouldn't it be ok to use atleast 1.75v, im getting the feeling this particular series of chips want voltage to operate correctly. I can only fear when operating on the very edge of your required voltage then unstability of Vcore fluctuation is sure cause that dreaded crash.

Before I brought the 3.2 I got put off it to see 2 reviewers only being able to overclock it to 3.45ghz with a voltage increase. Better count myself lucky then.

In all honesty, if you can get a 3.4 then do so. 3.4 is already a nice speed and surely intel would have to make certain its going to be stable at this, then overclocking you should atleast get somewhere. Forum threads ive read about this chip are saying positive things.

What did your stock heatsink look like?

[TC]

My retail hsf looks just like you described yours. I don't want to bother with air cooling since I'm going for max OC with the vapo. I tried up to 1.7 volts in the bios which gave me 1.68 acutal on my new ic7 max2. I'll try taking up to about 1.75 actual and see if it makes any difference. BTW a friend of mine here in the forum found this spec change notice from Intel which lists this chip as a D1 core with some packaging changes - namely the extra caps on the bottom, etc. Here's the pdf if you want to review. I may be dumping this chip for a 3.4 since I was hoping for 4.2+ when I decided to sell my prescott.

http://www.ocsetiteam.com/stuff/D0103696.pdf

Bumped it up to 1.75 and got this - but it rebooted after about 5 minutes of seti, so it's back to 4.0.

http://www.ocsetiteam.com/stuff/3.2.jpg

http://www.ocsetiteam.com/stuff/3.2bench.jpg

[AudiMan]

TC, what Vcore are you running at 24/7 @ 4 GHz?

[microfire]

This cpu sucks big time!
If the core voltage is not high enough then it ends up getting memory errors. Thats damm insane!
Makes games like UT drop back to the desktop with error messages.
The fresh hold of this happening is futher back than I expected. Even though I was running at 3.75ghz @ 1.675v without lock ups, its not stable even if I run 5:4 at loose timings, damm memory errors set in. Just been at 3.7ghz for the last 7 hours and had 5 errors. Looks like I will only to obtain 3.6ghz @ 1.675v for ture stable operation, and im not even sure of that until I test it for a good 24hours. At this voltage the cpu gets very hot, 62c full load with zalman7000cu on full speed, amb room temp 22c, I may as well brought a prescott cpu. Ive reseated the heatsink a couple of times now with different pastes and it is making proper contact.
I hate this cpu, this is the first cpu ive ever heard of that causes memory to get funky on ya.
I brought a dud :(

[stan03]

you might want to look into getting a different heatsink... 62 c is really hot.

[microfire]

The thermalright 900 heatpipe heatsinks are not here in NZ as I know of yet, and i didnt want to buy an old 900 heatsink with a high speed fan, reviews point out the performance between the 900 with quiet fan and zalman700cu were about the same.
I dont think I be able to get it much cooler even with a different heatsink, case flow is supplyed by 2x 80 intakes and 2x 80 outakes, room temp ranges between 24-30c, I think the real, the only time I see near 50-55c is if its about 18-20c in my room during the winter, problem is that a P4 at 3.5ghz+ is just using to much power, add a voltage increase on top of that and its cooking.
My ears have a problem with loud fans. Water would cost to much and at this point might not be a viable investment till prescott cooling solutions arrive.
Makes me wonder how hot the intel box cooler would be.

[TC]

Originally posted by AudiMan
TC, what Vcore are you running at 24/7 @ 4 GHz? So far 1.625, which equates to about 1.6 actual. I've been running seti and neverwinter for about an hour tonight without any problems yet.

[TC]

PS - the difference in load on my vapo compared to the prescott is amazing. With the prescott my vapo was holding about -15 under load at full speed. Now it's holding nearly -40 on the evap at 50% capacity with the 3.2@4.0.

[smoothmove]

I have the same chip. I am at 230x16=3679.7 @ 1.58 volts.

I can't f\do any faster than 238 FSB or it ends up crashing into windows or reboots during any program opening.

I was hoping to get to 4.0, but don't think so.

I have 1 gig of PC4000 HyperX 3-4-4-8 @ 1:1.

P4C800-E with a Radeon 9800XT PCI locked.

The chip is stable upto 3.71 Ghz- for me.

[microfire]

Are you on air?

I can get that far without lockups, but even if I back the timings off or reduce to 5:4 I end up with memory errors. It just failed 3.56ghz now im testing at 3.51ghz. Maybe ive got a danaged cpu.

[smoothmove]

on air using the retail fan.

[microfire]

Have you tryed running memtest86 on it to check for errors.
I can almost swear that my cpu is causing the errors and not the memory.
I use to run a 2.4b @ 3.1ghz / 172fsb, with memory at 4:5 = 215mhz, that was totally error free.
Man whats up with my cpu?
Should I try and return it? What else can I do. Usaully if its this hard, there is surely something wrong.

update.
Computer will no long post. I think the motherboard was dieing, and just packed it in?
I was getting errors in memtest86 at stock fsb.

[microfire]

Did I get a prescott heatsink or is everyone getting these with their 3.2C processors. Are these anygood on the northwoods?
Its the heatsink on the left is what I got:

http://www.legitreviews.com/Reviews/prescott_cooling_2.shtml

[Tedinde]

Originally posted by TC
PS - the difference in load on my vapo compared to the prescott is amazing. With the prescott my vapo was holding about -15 under load at full speed. Now it's holding nearly -40 on the evap at 50% capacity with the 3.2@4.0.

What are your real CPU temps. Chill control temps are a joke. Especially @ full cpu load.


-40???

[TC]

I can't really give you any real temps since I have an Abit board, but it's saying about -18C under load. The prescott on the AI7 was reading about 5C under load according to "uGuru"

[crotale]

Originally posted by microfire
Did I get a prescott heatsink or is everyone getting these with their 3.2C processors.

I got the same one with my 3.0C.

[microfire]

Finally I think ive found the limits of my 3.2c on air.
With 1.72v im able to hit 3.8ghz, only tested for 2 hours ut04 demo and 1 hour prime. Much high mhz and I can expect it to crash or reboot, also happens if I ad more voltage. 1.9v it will not post at all. That does not mean 3.8 will be stable but does give me a good idea of what its capable of.

Ive got a big problem and I can really put my finger on it. I can only run my ram memtest86 error free for hours at only 212mhz no pat and crap 2-3-3-7 timings :( Any thing above that gives me errors no matter what voltage I use, even 3.24v. Thought I can run the ram 1:1 upto 237mhz with voltage it still doesnt help down lower mhz.
I not that keen on running 5:4 and 170-190mhz, even here I get errors if I enable pat, voltage increase doesnt help, bandwidth is not good.
Ive almost tryed every setting possible, now im trying changing the modules around and then in the other slots, I hope it helps.
Ram useto run fine with my old 533cpu in this same board! at 415mhz and even 425mhz tight timings with the 4:5 setting, now that the 800cpu in is there, it doesnt want to know about it.
Is my ram just crap or did I just get a unlucky mobo

[microfire]

This cpu is looking damm good now despite using a zalman7000cu air cooling.
I had some away from the pc time today and ran prime95v23.8.1 at 3.752ghz for 18 hours and 30 mintues until I stopped it. Load temp was on 60c and vcore sat on ~1.63v

[Toaster Oven]

TC, was that -18C load temp for the 3.2c reported under uGuru? I thought that uGuru couldn't read negative temps? Or is there an update/fix that enables this?

[Andykat]

Hi Guys

I build a lot of systems based on highend p4s & i just had a bad experience with a 2,8 thats only just made 3.1!

So i looked on the shelf & checked the s spec on the 3.2s i have & both are the same as yours!

So i have just put one in my rig & started to ramp it up!

Doing 3.55 @ mo default vcore(1.55v)! if this is cpu is sitting @ 61c which is about 10c higher than my 2.4@3.3

ill keep u posted!