E6600, before and after vcore mod [Archive]

Epox4life

09-17-06, 11:19 PM

I wasn't going to post this here because there aren't that many phase cooling users on OCF, but I figured why not, I just want to show how voltage/vcore can effect and help your overclock, if your temps are good don't be afraid to use it.

This is a before and after comparison of how a vcore mod can help a LOT. All vapochill LS cooling.
E6600 retail, week 27.
P5B-Deluxe
2GB g.skill pc2-6400

I was completely stuck at 4.2ghz for Super PI 1M. vcore 1.64v real.
Before
http://star.walagata.com/w/fetuskiller/12.094small.JPG

Then I found CAL930's vmods, http://www.xtremesystems.org/forums/showthread.php?t=110121&page=5

And here are the results after a vcore mod, vcore 1.85-1.9v.
After
http://star.walagata.com/w/fetuskiller/11.343small.JPG

Just for fun heres a 32M run with 1.86vcore still using stock vapochill LS.
My board might be limiting me because the clock difference between the 32M run and the 1M run is very small.
After
http://star.walagata.com/w/fetuskiller/11.59small.JPG

hUMANbEATbOX

09-17-06, 11:37 PM

nicely done epox ;)

thassa lotta vcore. :D way to show that cpu who's boss. :beer:

maxxoverclocker

09-18-06, 01:26 AM

haha just don't kill that poor chip; xtremeoverclocker's been telling me about u :)

Phrenetical

09-18-06, 01:31 AM

Nice work epox, theres no way anyone should be running more then 1.65 unless your on phase so my question is

how did this help stability and voltage droop at 1.65, and where you able to push the cpu further on 1.65 with the mod or did you not check ?

edit: sorry i read that as vdroop mod, not vcore. Is there a droop mod for the p5b?

Epox4life

09-18-06, 01:48 AM

If you set the vcore to 1.7v in bios, it is really 1.64vcore then when you put both cores under 100% load the droop will go to 1.62v.
I did not do a vdroop mod, just higher vcore.

greenmaji

09-25-06, 03:16 AM

Baa.. you've done some Vdroop modage by now.. whats the 24/7 clocks like now. :D

Ross

09-25-06, 01:14 PM

Epox you freak! 1.85-1.90Vc??? :eek: I know retails run a bit cooler than ES, but I am still a bit surprised your LS had enough capacity to keep that voltage cool. Sick, but nice to know it will take that much Vc and not drop dead :D

Epox4life

09-25-06, 01:19 PM

Ross

09-25-06, 01:30 PM

Nice. That's about where mine is at with 1.66-1.67Vc. Either retails run a lot cooler then ES or my SS is due for some maintenance...maybe a little of both :shrug:

Gautam

09-25-06, 01:55 PM

Perhaps it has something to do with how your respective thermistors are calibrated. ;)

Ross

09-25-06, 02:16 PM

Epox4life

09-25-06, 02:20 PM

Haha, maybe, but evap temps aside, BIOS reads closer to 0C than I'd like and it doesn't seem to hold load like it did the first time I hit 4.8GHz and that was just a few weeks ago. Maybe it's a bit warmer in here or something. Either way, I'll need to give DI a shot soon. Maybe I'll try bumping up the Vc to 1.70V in the meantime...should at least be able to hold that for a 1M...or at least a screenie. I still have the pic of the 4.9GHz, but I'd rather have a validation :D
I still don't get how your doing these screen shots and validations so much higher then SPI 1M, I saw mine hit 4.74ghz before it locked up which is only 40mhz higher then my SPI 1M speed. Your 1M and 32M speeds are very far apart, while mine are close, these conroes are weird.

Ross

09-25-06, 02:40 PM

I just OC with Clockgen to the point where I can just get a validation. 1M can *maybe* go a bit higher, but part of the problem is that I don't really have enough capacity (especially for 32M). 32M is really limited because anything >1.60Vc at some point during a 32M bench will raise the temps too high and it will error/crash out. For validations, you just need enough cooling to hold the clock long enough...big diff from actually "working" on the clock ;)

Epox4life

09-28-06, 08:02 PM

Just got my second E6600, this time an OEM (open box) from newegg.
My other E6600 cannot come close to this FSB with the exact same setup, at any multiplier.
This chip also does 4.57ghz@1.84vcore SPI 1M stable on the vapochill, 70mhz better then my other chip.

E6600 OEM week 27, 1.84vcore @ 4.512ghz = vcore mod
G.Skill PC2-6400 HZ's @ 2.66vdimm = vdimm mod
P5B-Deluxe with HR-05 NB cooler

The cpu did come in a retail box though, although newegg sold it as OEM.

Picture of setup.
http://star.walagata.com/w/fetuskiller/P9280235.JPG

SPI 32M Result.
http://star.walagata.com/w/fetuskiller/untitled.PNG

Another example of vcore, use it guys.

dylskee

09-28-06, 09:06 PM

Nice chip Epox, but dayum that memory flies! :eek:

greenmaji

09-29-06, 11:13 AM

Nice chip Epox, but dayum that memory flies! :eek:

LOL.. Ask him what happened to windows ;)

Kids.. dont try that at home unless you have a well thought out backup plan for your OS :p
errors ahoy!!

But that 32M time is sick :drool:

No 24/7 clocks just yet.. asked to early.. up there.

Joe Camel

09-30-06, 02:04 PM

.

Epox,

what kind of temps are you getting @ those (crazy sweet) high V-core settings? (SS cooled not D-ice)

idle, load, BIOS what ever you have :)

Epox4life

09-30-06, 02:23 PM

Joe Camel

09-30-06, 02:31 PM

thought so... im playing around in the 1.7's and seeing -13 to -21 BIOS.

Chilly1 SS, lapped to copper IHS

so that sounds about right :D

thanks :beer:

i'll see if there is much/any gain for me into the 1.8's ... IIRC (my) P5W-DH has an over-volt limit or was it just a warning? ;)

Epox4life

09-30-06, 02:39 PM

It's just a warning, once you go into bios in the hardware monitor it will show the vcore in red font. :eek:

Ross

09-30-06, 03:57 PM

Not sure if these boards are the same since I haven't gone that high in Vc, but the red Vc warning may stop the boot with a warning. If it gets annoying after a few boots, just go into the H/W monitor screen and change the Vc reading to Ignore.

JC, did you see any gains by lapping? My ES runs so hot, I really can't bench 32M with more than ~ 1.60Vc actual, which is obvioulsy holding me back. It will get warm enough to crash almost no matter how low the CPU clock is at that Vc. 1M can go to about 1.66-1.67Vc since it's so much shorter. I was giving some not so serious thought to pulling the IHS completely, but don't think I'll ever get myself to do that so maybe just lapping will help?

Epox4life

09-30-06, 04:53 PM

Just lapped my E6600 that did 4.57ghz SPI 1M stable, will see if it helps at all.

Ross

09-30-06, 05:01 PM

Nice, I'll be watching for your results :)

Epox4life

09-30-06, 05:33 PM

First chip - 1.86-1.88vcore = -6C cpu in bios = -36C evap load > SPI 1M max 4.5ghz

Second chip - 1.84vcore = -2C cpu in bios = -36C evap load > SPI 1M max 4.57ghz

The second chip doesn't have as flat of an IHS, the edges are square instead of round.
AFTER Lapping

Second chip - 1.84core = -7C cpu in bios

Before lapping = 4.572ghz SPI 1M max
Edit* After lapping = 4.62ghz SPI 1M max

Ross

09-30-06, 05:44 PM

Hmmm. I dunno if that's enough of a gain for me to bother. Take the IHS off and let me know how that works :D

Epox4life

09-30-06, 05:47 PM

Hmmm. I dunno if that's enough of a gain for me to bother. Take the IHS off and let me know how that works :D
There's a Celeron D 341 on my keychain because of that.

5C difference at same vcore after lapping not good enough? Under load the difference will be even greater.

Just did a SPI 1M run at 4.62ghz @ 1.84vcore.

Ross

09-30-06, 05:58 PM

Probably, just not at my current laziness level ;) Same ambients as pre-lap? Keep me posted on your 32M difference after the lap...that would do more to convince me to pull it apart, or not :)

Epox4life

09-30-06, 06:05 PM

Probably, just not at my current laziness level ;) Same ambients as pre-lap? Keep me posted on your 32M difference after the lap...that would do more to convince me to pull it apart, or not :)
I basically killed my old install the SPI times are way off, lol. Now i'm making a new install.

Epox4life

09-30-06, 07:42 PM

Probably, just not at my current laziness level ;) Same ambients as pre-lap? Keep me posted on your 32M difference after the lap...that would do more to convince me to pull it apart, or not :)
My 32M time is limited by FSB on the vapochill 564fsb is the fastest I can run for 32M

Ross

09-30-06, 08:05 PM

There is one higher multi than 8x and I seriously doubt you are FSB limited there ;)

Epox4life

09-30-06, 08:17 PM

There is one higher multi than 8x and I seriously doubt you are FSB limited there ;)
I'll give 9x a try then.

Epox4life

09-30-06, 08:50 PM

For you Ross. 32M x9 results from the second E6600.

http://star.walagata.com/w/fetuskiller/11.36x9.PNG

Ross

09-30-06, 09:51 PM

Thanks Epox. What was your best 9x 32M clock prior to lapping?

I ran out and got just 5lbs. of DI so I am not up all night messing with it. In pretty short order it locked and wouldn't boot so I pulled the tube and ice had the CPU welded to the socket...guess I got a bit of an air leak somewhere and not enough DEG in the socket :rolleyes: I am drying it out now and trying to get it set back up before the little DI that I have evaporates into thin air. If I don't get back to it tonight, I'll probably get some more DI tomorrow ;)

Epox4life

09-30-06, 09:56 PM

Run SPI 1M first, I was planning to DICE this new E6600 (possible 4.8ghz for SPI 1M) monday so I could take #1 in SPI 1M for OCF, but since your doing it tonight I won't have a chance, lets see a 4.9ghz SPI 1M run.

I never tested this E6600 with 9x 32M before lapping, oops.

Ross

09-30-06, 10:02 PM

IF I get it running again tonight. The last time the SS got condesation in the socket it was 3 days before it would fire up again. You'll probably beat me anyway...or be close. I need lots more CPU clocks to make up for whatever the deal is with my slow(er) SPi times, especially 32M. I'd consider myself very lucky to crack off a 10.7, which you should be able to match/beat no problem.

Ah, that's what I was wondering, the before/after lap 32M, but thanks for running it anyway :) Gonna go check on how the drying is coming along...

Epox4life

09-30-06, 10:11 PM

I'm excited... :attn: Try for a 5ghz validation!

Epox4life

09-30-06, 10:47 PM

Another 32M run x9 multi.

http://star.walagata.com/w/fetuskiller/11.30x9.PNG

Ross

09-30-06, 11:01 PM

I'm excited... Try for a 5ghz validation!Oh, I got that on the first boot before it iced up...booted at 4.9GHz and clocked up to 5.00GHz :eek: It doesn't seem real happy about anything over 4.8GHz for benching though, even 1M :shrug: Only pulling 10.8s there :bang head It's up and running again though, so let me see what I can do...

Epox4life

09-30-06, 11:03 PM

Oh, I got that on the first boot before it iced up...booted at 4.9GHz and clocked up to 5.00GHz :eek: It doesn't seem real happy about anything over 4.8GHz for benching though, even 1M :shrug: Only pulling 10.8s there :bang head It's up and running again though, so let me see what I can do...
10.8secs at 4.8ghz something wrong.

Ross

10-01-06, 12:34 AM

I need lots more CPU clocks to make up for whatever the deal is with my slow(er) SPi timesI still dunno what the deal is. If it were any other person, it would be a 10.6 1M probably :rolleyes: I'll be out of DI before I get to any 32M I think, but at best I'd be very lucky to get 11:40 anyway, so no big loss there (except for our team average). I guess the only "good" thing so far is the 5+GHz validation :shrug:

Gautam

10-01-06, 12:46 AM

Something is wrong. :confused: Perhaps you're teetering on the edge of the stability there? That often hurts times.

Anyways props on the suicide, 5 GHz = teh sex :cool:

Ross

10-01-06, 12:59 AM

Maybe, but my SPi's have been on the slow side pretty much always. My 11.0x was at the same clock JC got 10.70 at on his DH IIRC...big diff. What bothers me is why mine is so much slower than Epox since we're on the same board. At this rate, I'll need 4.9GHz for 10.6x :rolleyes: If anyone has any ideas, I am all ears :shrug:

I'll post the validation when I am done for the night...

Epox4life

10-01-06, 01:16 AM

I just want to see your SPI 1M run, I don't know what the problem could be, try reflashing your bios maybe?

Ross

10-01-06, 02:38 AM

Oh well...this might have been a 10.6 if I had time to try 4-4-4-8 :shrug: Either way, pretty lousy time for the clock. 706 BIOS was installed yesterday and OS was installed today. It's been a bit slow like this since the beginning and I am not sure if it's the CPU, NB, mem or what, but apparently I am either missing something in plain sight or it's something I can't do anything about.

Too bad though because these kind of clocks at 1.70Vc on DI is nuts :eek: I wish I had the balls to try 1.80+ like you ;) The board is only good for 570 benchable FSB anyway (580FSB suicide)...best I could possibly hope for is 5130MHz, so I am not too disappointed in the clock...just need to do something about the SPi times :shrug:

5054MHz validation: http://valid.x86-secret.com/show_oc.php?id=126121

http://i10.photobucket.com/albums/a125/boostedevo/dice/5050.png

http://i10.photobucket.com/albums/a125/boostedevo/dice/1m_10718.png

Epox4life

10-01-06, 02:42 AM

sooo close to #1 SPI 1M time on OCF, thats some serious mhz though.

Here is a screen of my memset. 2.64vdimm.
http://star.walagata.com/w/fetuskiller/memsettings.PNG

Ross

10-01-06, 02:55 AM

Yeah, it's really bugging me (time vs. clock). It should easily be 10.6 even with the loose secondaries. I can run 9s in the write->precharge and wr->read delays, 4 in tRAS and 20RFC with 2.50Vddr, but I left them at what I was booting with. Might have only been a 10.68 at best if I tightened them all up anyway, LOL. I wish I knew what's up though because it's going to drive me nuts :bang head

Epox4life

10-01-06, 03:00 AM

You really need to test the cpu in a different motherboard or try another cpu.

Ross

10-01-06, 03:07 AM

Yeah, I just got my RMA DQ6 back this week. I should try it in there, but it certainly won't be on the SS or DI...not insulating that board to turn around and sell it next week. Of course, for all I know, it's one that does 600FSB ;)

Epox4life

10-01-06, 03:10 AM

Ross

10-01-06, 03:16 AM

Hehe...yeah with my luck, it will be a 600FSB board and will still manage to be slower than just about everyone else in SPi ;)

Epox4life

10-01-06, 03:24 AM

Hehe...yeah with my luck, it will be a 600FSB board and will still manage to be slower than just about everyone else in SPi ;)
Don't you have like an E6400 or something you can pop in the P5B-D to see if the 32M times are on point?

Ross

10-01-06, 03:35 AM

I did, but I sold it to a friend of mine a couple weeks ago. I *almost* sold him this 6600 instead :eek: I'm not really in the mood to pull the 6600 out of the P5B and try it in the DQ6 because it's gooped up like crazy now (and no more condensation either), but I guess that would be the only way to make the comparison. Bah, maybe I'll try it tomorrow...

dominick32

10-01-06, 08:23 AM

Let me just say this is a great thread!!! :beer:
Epox and Ross, these clocks are incredible. What grit paper did you use for the lapping and what was your process for smoothing out imperf's?

Ross

10-01-06, 01:05 PM

Thanks dom. I might need to lap for a shot at the E6600 WR (5130MHz) since SPi times aren't cooperating ;)

Gautam

10-01-06, 01:12 PM

I think that WR's in the bag for sure. Pump in some more v's and it should be np. :D

About the Pi times this might be obvious, but how's your windows installation? It's only with Conroe, but it seems like for me dramatic changes in speed kills them. So far I've just had corruption, but others have had performance go down the toilet too. A reinstall might be all you need to get the pi times back to speed.

Ross

10-01-06, 01:30 PM

I dunno about in the bag, but I'll sure give it a shot. I think I might lap it today and if that doesn't help very much, maybe I'll work up the nerve to pull the IHS, but I doubt it ;) I don't know if the vdroop mod would help at all either (for benching)....it only drops about .025V on load.

XP was reinstalled yesterday and 706 BIOS the day before. I should go back to 507 BIOS...definitely good for ~.1s faster (max) in 1M from past experience. Somehow I managed to forget all about changing the BIOS change last night, but I guess I was a little flustered after the condensation issue. I guess it wouldn't hurt to check out any new chipset drivers too...still using the ones that came on the mobo CD.

fritzman

10-01-06, 02:26 PM

Once you do all that, you'll be obligated to update the timings in that very fine avatar of yours.

You're doing a great job = inspiring... Sounds a bit lame, but it's actually true. There are a handful of guys here (several already appearing in this thread) who are consistently at the top, and when you pool/share info (as you guys do) willingly with each other and members generally, it makes the pathway for us plebs that hang round near the table, much more understandable.

From me... thanks!

Ross

10-01-06, 03:38 PM

@fritz: :beer:

I am lapping away and had to give my arm a rest. Mine was high in the middle, so I don't think it's going to help much/at all except for maybe making it .75mm thinner :) At this point, I definitely think my only temp decreases would come from removing the IHS completely, but I'll finish this up and give it a shot first. I am not sure if I'll get more DI today or not, but maybe I'll toss it in the DQ6 board real quick when I am done and compare SPi's.