Thats funny you mention the vtt under volting, I was looking at ET6 and noticed that my vtt read 1.30. I will change that and give it another try.Quote:
Originally Posted by Hoss331
Printable View
Thats funny you mention the vtt under volting, I was looking at ET6 and noticed that my vtt read 1.30. I will change that and give it another try.Quote:
Originally Posted by Hoss331
I usually do 8-12 hours depending, I just figured at such a low oc (470X9) it should be able to run prime for at least 24 hours.Quote:
Originally Posted by andressergio
10 hours (i.e. enough for a complete loop of all FFT sizes in blend) is enough for me.
Still having a fair bit of trouble around the 480k/20k test area (around 4-5 hours into blend), with either hangs or rounding errors. I'm testing at 489x9/1.37v with memory at 1223 on 2.50a divider. It seems to like 1.28vtt the most, any lower and the board gets stuck at POST after a reboot, any higher, and blend fails sooner. MCH is at 1.54v and Tread is 8. Vdimm is set to 2.220v, although both do just as well with less. Have tried various voltages skews & ref changes, but still can't get through a complete error-free loop at this speed. CPU temps dont go beyond 65°C so I'm starting to suspect the board, but if anyone can offer any hints I'm all ears!
You shouldnt need anywhere near that kind of mch voltage for that oc, more likely the 1.40 area. Have you tried running it using 2.40B instead of 2.50A?
yes agree with you Cliff no way such MCH 1,4 will be good but he is using 2,50A wich is buggy and a hell on the MCHQuote:
Originally Posted by Hoss331
Cheers
Sergio
I remember you having problems with 2.50A thats why I asked if hes tried using 2.40B instead. For one reason or other my board has no issues with 2.50A nor does it seem to need alot of extra voltage, 1.40mch for 4.5/1250. 2.40B on the other hand has always had issues with how much mch voltage it can handle at a given fsb before it will bsod or just not boot. Where your able to run 1.46+mch for 500x9 I cant go over 1.40 on 2.40B without have problems, 1.42 is kinda usable but 1.44+ will crash instantly or not boot at all. Theres a narrow usable window on how much mch voltage is needed to be stable yet not crash from having to much voltage and this scales with fsb, as it increases so can mch as it deceases so must mch or it becomes unstable. This is using the red slots, if I use the yellows the max usable mch voltage drops by about 0.02-0.04 (to about 1.36-38 at 500fsb).
Under 470 FSB it is very stable for me, however over that it is very buggy. 2.40B is always more stable at higher FSB.Quote:
Originally Posted by andressergio
yes i really dont have a problem just need lots of volts that dont justify for just 50mhz more...but remember that u have the blades...the gskill combo + that 2,50A diver is hardQuote:
Originally Posted by Hoss331
Cheers
Sergio
Yea it might just be a compatibility issue with the gskills as my cellshocks also dont have any problems with 2.50A. :shrug:
Thanks for the helpful responses. I wasn't aware of the 'limits' you've experienced when bumping vMCH at certain speeds. I've not experienced these myself, but assumed 2.50A with PL8 would require more volts. I'm now testing at 1.4v, but with memory in yellow slots so I'll get back to you all. I'lll take the info on board about the mch voltages, but these odd voltage limitations some of you appear to suffer don't seem like an issue with me, but then again my n/b is watercooled which may/may not make the difference. Perhaps it's a rev1.6 thing?
I'm hoping to finally have my EP45 extreme tomorrow, and I've got a few more days off to see what it can do. If it does any better, this will end up being my 24/7 board.
I've been toying with an E7500 (11x multi) for the past day, no records broken but it certainly could take the voltage. I managed 4.95GHz with 1.88v on H2O, but was only benchable at <4.7GHz/1.8v. Temps never went beyond 65°C and coolant didn't even get warm, but I got a bit scared after that. I thought any higher without anything cooler could just fry the chip, particluarly as most others were benching with ss/dice etc with those volts. I'll get thoose pic up on hwbot soon, only worth a couple of points though, if that!
My problem with mch and 2.40B are probably just an issue with my particular board, you may very well not have an issue with it or 2.50A. My nb is also water cooled and when I was still benching with 2.40B on the cold nights it would get even more sensitive to mch voltage. Do you know if the extreme has been moded yet? It has no LLC and apparently the vdroop is pretty substantial in unmoded form.
In-Fluence Do you really get 4.2GHz with that voltage ? and just how ? Man mine needs so much juice it at funny .
1.4 Volts For 4.2GHz . Man would be great to get them lower .
http://img94.imageshack.us/img94/1131/newocv.jpg
Quote:
Originally Posted by trickson
Asus motherboards are not so great at quad oc ... u need a Gigabyte EP45-UD3R/P for high clocks with low voltage
Some chips will just clock better than others mate, and as your chip had a VID of 1.25. I would say that's 0.05v higher than mine so the difference in voltage makes sense. FYI the Asus MIIF actually got all the way up to 4.3GHz and it acheived that with lower volts than my UD3P, probably because of the finer GTL tuning options. It's a shame because that all I reckon it needed were 50ps steps in the CPU/NB skews and 500fsb may have been attainable.
Anyhoo, my EP45-Extreme is here at last - not happy about the seller leaving the giant coper heatsink rattling around inside the package with the board, but there's only a couple of scuffs and hopefully there'll be no damage to the components. Will no doubt post back with results later :)
+1000000000000000000000000000000000000000000000000 00000 for thatQuote:
Originally Posted by kagula323
up giga up !!! :D
Been tinkering with the board today, finally getting stableish at the same volts, unfortunately this board vdroops like nothing I've seen before (like you mentioned Hoss), so I'm popping down to maplin 1st thing tomorrow for a 5k pot. I'm looking at .1v droop from idle to full P95 load! This means I've had to set ~1.52v in bios to try for any stability attempts at 500x9 - idle temps were acceptable, but still uncomfortable for 24/7. Memory overvolts by about .04v too, but runs nicely at 1200/2.03v :D Hopefully, with the mod, I can begin to finalise my seemingly endless quest for a 4.5ghz/1200MHz rig and setlle down to using it!
*edit - link to my post re: vmod and results ;)
http://www.xtremesystems.org/forums/...postcount=6183
Hey everyone.. I was here last week still wishing I could get this ep45.. but I now might be ready to join the 4.5 GHz club. Lol
Fry's electronics I found a open box EP45-UD3P for $79 just missing the I/O shield. I'm seriously thinking about getting it. Thoughts? I no longer have SLI . My GTX 480 will be here tomorrow.
I have a good early batch 1.2500 vid Q9650 doing 4.2 with a nf200 board stuck at 1.35 fsb/vtt voltage and using way more than 1.304v for 4.0 ghz like sig says. It boots with that. But needs much more for stability.. I think I could get much higher with the ud3p. Now might be my chance.
Does anyone know how well my RAM will run in this board? I have a 2x2GB OCZ Blade 1066 MHz kit I could try instead. They are NOT LV tho.
Should I snag this board?
I would certainly do it. I had one setup a while back with a 9650 and it was much better then any board to that date for 775 and I have tried them all. Fantastic results with much higher stable fsb clocks for both the 9650 and 9550.
It'll take your quad a fair bit further and possibly with lower volts (and your vcore is not that bad at the moment). UD3P boards are the best you can get for o/cing them, I'd go for it if I were you, they're becoming very hard to get now. If you're willing to fork out a bit more cash, the usb3 version might be up your street and more easily available to the US.
My EP45 Extreme is finally holding my CPU stable at 4.5GHz but it needs a bump in vcore (i.e. 1.475 in bios / 1.45 real). The vdroop mod has really paid off and now it seems to have done better than my UD3P, but this large bump in vcore required to get stable has also knocked temps into the high 70's during a few hours of small ffts - now looking to try and lower it a notch or two and getting a rad better suited to my low-noise fans.
http://i194.photobucket.com/albums/z...Q965045GHz.jpg
Hey guys just a heads up that my golden Q9650 is up for sale:
http://www.xtremesystems.org/forums/...d.php?t=249851
http://img683.imageshack.us/img683/9292/3580too.th.jpg...http://img508.imageshack.us/img508/4335/jor3lbr1.th.jpg...http://img31.imageshack.us/img31/124/jor3lbr2.th.jpg1
Time to move onto more chips isn't it? :DQuote:
Originally Posted by Jor3lBR
Hope it gets a good home!
Jor3lBR,man i cant believe it.wish i still had me UD3P
well i have my own Q9550 that can do 4.25 on 1.264 volts but Seriously if i had the money and didnt had a Q9550 i Surely bought it its a verry nice cpu lets hope it comes into good hands :)
It should be fun to use Jor3's chip, don't expect the same clocks he got though.
It will indeed take quite a push to get it back to those record speeds. good luck with it.
having a small problem with getting my setup stable again
I just replaced a broken Rosewill 950w with a Corsair HX750 (I wanted the HX850, but my local microcenter didn't have any in stock) I am constantly getting freezes in Prime 95 blend. I have no stability at any of my other profiles as well.
Also I replaced my lapped TRUE for a Venomous X, but I hardly doubt that would affect any related to my overclock. The 12V rail seems rock solid at 12.3, the 3.3V rail is at 3.25, and the 5V is at 5.1 under full prime 95 blend load.
Any idea's?
Sounds like a ram or MCH problem. Give a bump to the MCH, if that fails try and adjust the MCH ref's.Quote:
Originally Posted by l0ud_sil3nc3
What part of prime blend is failing, large FFTs I assume? Curious you should mention you got a hx750 as I lost a little stability on my old Q6600 about a year back when I changed over to my hx850. Although the PSU runs nice & cool, I'm starting to get suspicious now :s
Quote:
Originally Posted by In-Fluence
it freezes within seconds, maybe thirty:shrug:
I am thinking I may return this ps I just a need a few more days to test.
MCH reference along with VTT and MCH voltages are probably the cause mate, I wouldn't give up just yet. The aforementioned are crucial to stability on this board, and could result in a bootable, but not 'stressable' system. You might even find that your old PSU required the board to give higher voltages, and you might need to compensate/re-tune.
List your settings and perhaps we can offer some more suggestions ;)
A quick bench session http://valid.canardpc.com/show_oc.php?id=1151229
That's my baby! Told ya it's simple the best :DQuote:
Originally Posted by NoGuru
I switched from a HX1000 to a HX850 and I'm still stable with the same bios settings, so it was prolly something else in your case (unless you got a faulty unit ofcourse). My 5870 actually clocked slightly higher at the same volts AFAIK. Not scientificly tested though. Never did a before/after test, but it felt that way. Might have been a placebo too :p: Overall I'm still happy with my HX850 thoughQuote:
Originally Posted by In-Fluence
Yeah it's great! Maybe it thinks it's a duel core, it sure clocks like one.Quote:
Originally Posted by Jor3lBR
I can't thank you enough.:up:
Congrats NoGuru i was waiting for a bench :up:Quote:
Originally Posted by NoGuru
Jor you are dad for us :toast:Quote:
Originally Posted by Jor3lBR
i have the same amazing PSU the HX850 :up:Quote:
Originally Posted by Musho
I will have more time next week to really put a hurt on it.:up:Quote:
Originally Posted by andressergio
NoGuru,amazing OC.i always thought jor's chip was in a different league
than the rest of ours.cant wait to see some more of your results:up:by the look's
of it,the chip is in great hands.im glad you got it,at least you'll push it to it's max
Thanks radaja. I will push it, but I don't have Jar3's skills yet to push it to the max. I am going to try and save for some WC. I'm a bit clueless on what to get, but I am doing some research now.Quote:
Originally Posted by radaja
awesome.WCing is the way to go with a gem like that.
yea i was confused by what to get but it's not really that hard to figure
out just waht you need.i asked members here for advice and people like
zsamz and andressergio,helped me out quite a bit.so feel free to
ask questions
you got us mate ;) you need to cool it thats the first thing to do, and well...Quote:
Originally Posted by NoGuru
Thanks radaja, so many ppl helped at the start time to return that :up:Quote:
Originally Posted by radaja
Quote:
Originally Posted by radaja
Thanks guys. I already bought a D-TEK waterblock. It's a Fuzion V1, but for $20 IQuote:
Originally Posted by andressergio
could not pass it up.:yepp:
didnt know Jor was a geezer :rofl:Quote:
Originally Posted by andressergio
lots of love in this page:up:
not a gay thang:rofl:
grats Noguru on the cherry:up:
i still hate gigabut:yepp:
but i like the udrX:D
and there goes zsamz :rofl:Quote:
Originally Posted by zsamz_
Jor pardon this man thats so lost :rofl: he is searching the "bu**t" on a mobo :rofl:
what lolQuote:
Originally Posted by andressergio
did i say somethin wrong:p:
you callin him daddy:p:
gigabut didnt make a good board till the ud3X series
took em over 10 years of lemons :rofl:
my first gigabut p45 craped out but that was expected :ROTF:
:rofl::rofl::rofl:Quote:
Originally Posted by zsamz_
Finally got around to doing so more dice testing and got a new best for spi 1m. I was planning on doing a max speed run also but didnt get around to it, hopefully tomorrow night.
http://i131.photobucket.com/albums/p...1/921Image.png
bravo Hoss331
:clap::clap:
Hoss331,thats some awesome quadcore fsb action for sure.i couldnt even get my last two
UD3P's to do fsb above 550 using a dualcore.simply put,WOW
Hoss, great score. What ram are you using?
A 5Ghz+ SPI run, Very nicely done Hoss!
I've persuaded an old friend to ship me a USB3 board now which will be my last 775 board for certain. I'm not hoping for more success in benching, I just love the colours they've used and know it'll look great in my case. Keep pushing hoss, how's your max CPUz coming along? :D
*edit scratch that - just seen it 5.1+GHz - go Hoss!
Thanks :up:Quote:
Originally Posted by E-mil
I cant complain its my original P board and it still puts up the FSB numbers, hopefully my ddr3 board will also when it gets its chance.Quote:
Originally Posted by radaja
The ram is radajas old PC8000 C4 Cellshocks (D9s).Quote:
Originally Posted by NoGuru
I actually had a 8.907 run done at 5.103 but it locked on me just before I could click save in paint. :( I was low on ice so Ill get it next go around and try for a new max validation, with the way its running im thinking a 5.2 is within reach. :yepp:Quote:
Originally Posted by In-Fluence
The dreaded lock up before the save. I have faith you will get it next time.
After reading through many of the posts on this thread I realized that I wasn't just dreaming on getting my Q9650 above 3.6GHz...I previously had it running on an MSI P43-Neo3F, but nothing I did would let me go above 3.7. After reading many of the posts here, I knew I should be able to get more out of it and I became convinced that I was being limited by the low native bus speeds of my board and DRAM (PC2-8500)
After a quick trick to the local Fry's, I picked up a Gigabyte EP45T-USB3P and some brand new DDR3, with the push of a button I was up to 3.8GHz, and with a little extra on the CPU voltage I run 4.05GHz with almost no effort. This PROC is AWSOME!
I'm now Prime95 stable at 3.8 for daily use with ocassional desired to see it at 4.05GHz - hav'n't completley stabilized the 4GHz setting yet as I get a spontaneous reboot when running Prime but generally very happy with this chip - thanks all for your various posts and input that convinced me that under the right environment my chip was capable of much more than I was getting.
Yeah with that board you can get this chip pretty high. Post your settings and we can help you tweak them. Oh, and welcome :welcome:Quote:
Originally Posted by b1tphr34k
I am having a problem with my USB3P, I can't set the command rate to anything but 2T. Every time I try to change it, in windows it is always 2T.
I even tried using Memset to change it and it will not allow me to change it as well. The option is available in the drop box but when I choose it, it still comes up as 2T.
I have tried looking for a beta bios or any other bios other than F1 to try, because it seems the memory dividers are botched on this board, only 3.20B works for me.
i know on intel chipsets with ddr2 1t is not allowed,but i just checked through the UD3P thread
and still cant find any ddr3 boards using the 1t.all their SS show 2t.but i will keep looking for you
EDIT:
according to 3oh6 his board had the option but wouldnt run 1t either.
P45 and DDR3??
see that sucks kinda makes me wish i kept my UD3P:shrug:Quote:
Originally Posted by radaja
I still can't get this memory to run cas6 @ 1600, much less even post at 2000.
I have had my MCH voltages up to 1.46 and still no dice on any of the 4.00 dividers.
you gonna need @ least 1.7vQuote:
Originally Posted by l0ud_sil3nc3
and no way you runnin those cas6 if they are bbse
Quote:
Originally Posted by zsamz_
I assume I would have to take the heat spreader off to find out?
Hell they won't even run cas 7, I am only getting away with running cas 8.
I have this kit KHX2000C9AD3T1K2/4GX
I ended up buying three of them so I could do 6GB in my X58 box and 4GB on the P45 rig plus I have one extra stick.
I can still return them, I want at least 1600 cas 7 and if I can't get them to run that then they gotta go.
i had a few sets of bbse and i had tough time with them on rex 1 set toped off 1600 cas7 other sets 1700 cas8
the 2200 cas7 pi toped off 1920 cas6:up:
you need microns or hypers
there are a few sets with micron in sale section;)
i would grab these hypers soooo fast
http://www.xtremesystems.org/forums/...d.php?t=250371
but i dont think he shippin to canada
Quote:
Originally Posted by l0ud_sil3nc3
unfortunately my post count is too low, so I can't see that forum.Quote:
Originally Posted by zsamz_
i have horrible luck with getting low VID chips
http://i44.tinypic.com/2816d5k.png
That's not always a bad thing, lower VID sometimes equals higher heat.
Thanks! so I decided to go ahead and try again - this time I made it a point to back off the memory and only push the CPU.Quote:
Originally Posted by NoGuru
So far Prime95 table...but definatley appreciate any insight that might tighten things up a bit. Basically, I started with ET6 to 4.05, added PCI=100MHz, upped CPU V1.34, backed off DRAM to standard (2.66D) @ 1200MHz and left everything else auto. I'm not so sure she should be cooking at 72-73 only 10 minutes, but open to suggestion :) after reading through much of this it seems the MCH and VTT are where a lot of the sweet spots lie? but I'm not sure what I should look for there.
Thanks for your help!
http://i209.photobucket.com/albums/b...dsadler/OC.png
this one does run really cool though 67c loaded @ 1.44Quote:
Originally Posted by NoGuru
Your PLL and VTT are off the charts. Set PLL to 1.50 and set VTT manually, about 1.32 just inch your way up. Usually it stays just below vcore.Quote:
Originally Posted by b1tphr34k
:up:Quote:
Originally Posted by l0ud_sil3nc3
JorB3's ram is awesome:yepp:
I'm still working on tunning things up a bit, but I thought I'd share progress and see if anyone could offer additional ideas that might help stabalize.
The system runs great, but still BSOD/Reboot under Prime95 within 10 minutes. Temps have cooled off several degrees C, and I gained an additional 100Pts in Passmark score by boosting my RAM just need to stabalize it under load and things will be great.
Here's the BIOS code I'm running
Here is a shot under loadCode:GIGABYTE GA-EP45T_USB3P
[u]MB Intelligent Tweaker(M.I.T.)
Robust Graphics Booster ...............: AUTO
CPU Clock Ratio ..........................: 9X
Fine CPU Clock Ratio.....................:0.0
CPU Frequency ...........................: 4.05GHz GHz
Clock Chip Control
Standard Clock Control
CPU Host Clock Control..................: [Enabled] []
CPU Host Frequency (Mhz) ............:450Mhz
PCI Express Frequency (Mhz) .........:100Mhz
C.I.A. 2.............................:[Disabled]
Advanced Clock Control [Press Enter]
CPU Clock Drive...........................:800 mv
PCI Express Clock Drive...................: 900mv
CPU Clock Skew (ps).......................: 0ps
MCH Clock Skew (ps).......................: 0ps
DRAM Performance Control
Performance Enhance....................: [Standard]
Extreme Memory Profile (X.M.P.)........: Auto
(G)MCH Frequency Latch.................: [Auto]
System Memory Multiplier ..............:3.33D
Memory Frequency (Mhz) ..............: 1500Mhz
DRAM Timing Selectable ................: [Manual] []
Standard Timing Control
CAS Latency Time...............................: 8
tRCD ..........................................: 8
tRP...........................................: 8
tRAS...........................................: 24
Advanced Timing Control [Press Enter]
tRRD...........................................:AUTO
tWTR...........................................: AUTO
tWR............................................: AUTO
tRFC...........................................: AUTO
tRTP...........................................: AUTO
Command Rate (CMD) ............................: AUTO
Channel A
Channel A Timings Settings [Press Enter]
Static tRead Value......................:Auto
tRD Phase0 Adjustment...................:Auto
tRD Phase1 Adjustment...................:Auto
tRD Phase2 Adjustment .................:Auto
tRD Phase3 Adjustment..................:Auto
Trd2rd(Different Rank)..................:Auto
Twr2wr(Different Rank)..................:Auto
Twr2rd(Different Rank)..................:Auto
Trd2wr(Same/Diff Rank)..................:Auto
Dimm1 Clock Skew Control................:Auto ps
Dimm2 Clock Skew Control................:Auto ps
Channel A Driving Settings [Press Enter]
Driving Strength Profile:[AUTO]
Data Driving Pull-Up Level.............:Auto
Cmd Driving Pull-Up Level..............:Auto
Ctrl Driving Pull-Up Level.............:Auto
Clk Driving Pull-Up Level..............:Auto
Data Driving Pull-Down Level...........:Auto
Cmd Driving Pull-Down Level............:Auto
Ctrl Driving Pull-Down Level...........:Auto
Clk Driving Pull-Down Level............:Auto
Channel B
Channel B Timings Settings [Press Enter]
Static tRead Value......................:Auto
tRD Phase0 Adjustmen...................:Auto
tRD Phase1 Adjustment...................:Auto
tRD Phase2 Adjustment .................:Auto
tRD Phase3 Adjustment..................:Auto
Trd2rd(Different Rank)..................:Auto
Twr2wr(Different Rank)..................:Auto
Twr2rd(Different Rank)..................:Auto
Trd2wr(Same/Diff Rank)..................:Auto
Dimm1 Clock Skew Control................:Auto ps
Dimm2 Clock Skew Control................:Auto ps
Channel B Driving Settings [Press Enter]
Driving Strength Profile:[AUTO]
Data Driving Pull-Up Level.............:Auto
Cmd Driving Pull-Up Level..............:Auto
Ctrl Driving Pull-Up Level.............:Auto
Clk Driving Pull-Up Level..............:Auto
Data Driving Pull-Down Level...........:Auto
Cmd Driving Pull-Down Level............:Auto
Ctrl Driving Pull-Down Level...........:Auto
Clk Driving Pull-Down Level............:Auto
{U]Motherboard Voltage Control{/U]
CPU Vcore . ...........................:1.32500v
CPU Termination ...... 1.200V*...........: 1.3000v
CPU PLL .........1.500V*...........: 1.5000v
MCH/ICH
MCH Core ..........1.100V............: AUTO
MCH Reference.. ... . .0.800V............; AUTO
ICH I/O .... ......1.550V............: AUTO
DRAM
DRAM Voltage ........1.500V............: 1.6v
DRAM Termination ... ..0.900V............: AUTO
Channel A Data Vref ...0.75V.............: AUTO
Channel B Data Vref ...0.750V............: AUTO
Channel A Address Vref 0.75V.............: AUTO
Channel B Address Vref 0.750V............: AUTO
[b]{U]Advanced Settings[b]{U]
Limit CPUID Max. to 3.....................: [Disabled]
No-Execute Memory Protect............: [Enabled]
CPU Enhanced Halt (C1E)................: [Disabled]
C2/C2E State Support....................: [Disabled]
x C4/C4E State Support..................: [Disabled]
CPU Thermal Monitor 2(TM2) ...........: [Enabled]
CPU EIST Function.........................: [Disabled]
Virtualization Technology................: [Disabled]
Note: * refers to 45nm CPU
http://i209.photobucket.com/albums/b.../Capture-1.png
For starters dont bother with the 3.33D, it doesnt run at its rated speed, instead use 3.20B. If you want to see the difference check the bandwidth on each.
Thanks HOSS, I'll have to check that out tonight - I actually made some other adjustments late last night to MCH, and VTT/Vcore -the result was slightly lower temps and faster Passmark score (by about 100pts)
I think I'm getting closer - I really wanted to try to get the RAM as close to 1600 without stressing it but it seems pretty touchy.
Quote:
Originally Posted by b1tphr34k
First turn off cpu thermal protection, then enable vdrop correction, because to me it looks like its not turned on, since you have 1.325 Vcore set and you are seeing 1.23 at cpuz.
Next set your vtt to 1.2 and then set your cpu ref to 0.785 and the mch ref to 0.800, and then go back and set vtt to desired voltage i.e. 1.30 as you have selected.
Give that a try now:up:
Took sil3nc3's advice and made the necessary adjustments. didn't want to take a chance on lowering my RAM settings just yet - I actually want to get that all the way to 1600MHz at some point.
I'm running Prime as we speak and getting much better results - OH - I regreased by cooler as well (used cheap thermal grease last time, and found my bottle of Zahlman's ceramic) makes a difference for sure.
I'm 5-7 C cooler, OH and I lowered the VCore, with the droop I noticed it was actually running 4GHz loaded using as little as 1.248 CPUZ, even though I had set it to 1.34BIOSv (so I lowered it to 1.275BIOSv with calibration (prevent droop) and its running Strong at 4GHz and 1.248vCPUZ
Thanks all for your help! If anyone can recommend anything else I could try to fine tune it that would be great. I already tried lowering the voltage in small increments - 1.275 seems to be the lowest it will run...what about some of those other values - should I now try adjusting them higher or lower to gauge impact to performance, temps, etc?
Thanks for all the help!:D
Going to 3.20B from 3.33D is by no means lowering your ram settings. Like I said earlier it doesnt run at its rated speed, its bugged and really only a 3.0, so at 450fsb your ram is running at 1350 not 1500. Heres a couple everest benchs for you. :up:
3.33D C5
http://i131.photobucket.com/albums/p...81/333DPL7.png
3.20B C6
http://i131.photobucket.com/albums/p...81/320BPL7.png
Do you really think I can run cas5 if my ram is actually at 1500 and still have lower bandwidth than 1440 cas6. ;)
thats a nice:clap:Quote:
Originally Posted by Hoss331
demonstrates the point real well
That makes perfect sense - sorry I didn't pick it up the first time.
i'll try that out t'night - thank you for taking the time to explain, I gues I just didn't completley understand what you were trying to say the first time around.
Quote:
Originally Posted by Hoss331
Are there any specific settings Imight want to work with (DRam voltage, etc) when I try to make that change I get boot failure fron windows and frequent BIOS lockup before POST completes, it was hard just to get back in and change it back before it locked up - really didn't want to have to redo all the tweaking (havn't been great about saving it as I go - changed that now of course)
Thanks for all you guys help
EDIT - on a separate note - I seem to be performing very well and Prime Stable for at least 30 minutes (yeah I know its not like 2 hours or 24 hours as some like to run - but I stopped it when thermals on one of the core's crept up to 74C, am I being a little over protective of my chip or do I need to look for ways to improve my thermals before continuing to attempt longer prime runs?
What ram kit are you using? If your only at 1.6 vdimm now you may need to bump it up some. As for your temps I wouldnt worry about them to much at this point as long as they dont go much higher during prime.
RAM KIT KHX1600C8D3K2/4GQuote:
Originally Posted by Hoss331
rated up to 1.65v I believe, I tried 1.64 with no luck, next option was 1.67 but I wasn't sure how high I should take it over the rating
Do you still have your mch voltage set to auto? If yes than set it to 1.32v and also disable xmp.
sweet thanks! I'll give that a shot...I got my thermals way down finally - in the middle of a prime run at only 63C right now...which is much better (new grease, new fan, and needed to turn one of my intakes into an exsaust since my rear exsaust became an intake (for the Corsair radiator)Quote:
Originally Posted by Hoss331
I'm hoping this will run for at least 2 hours, then I'm off to play with the ram again...
My new batch
vid 1.1875v
http://img338.imageshack.us/img338/9...3524552935.jpg
Sweeeeet what's the batch #
on the screen
batch #L004B845 packdate 04/09/10
dam its nice to see they are making nice chips in 2010, thats loads better than my Xeon.
very nice pvhk
Great VID but Super Pi = Piece Of Cake.Quote:
Originally Posted by pvhk
Give us some Linpack / Prime95 Blend screens.
1.2VID
http://img188.imageshack.us/img188/1...voltage.th.jpg
http://img706.imageshack.us/img706/9...oltagey.th.jpg
impressive fsb form x38 mb, how much vNB u use ?:P
my other q9650 vid 1.2375v (1.10v=wrong value)
on DFI UT P45
http://img100.imageshack.us/img100/7...2662436259.jpg
http://img697.imageshack.us/img697/9...5963996.th.jpg
1.69VQuote:
Originally Posted by papatsonis
@pvhk:
What is the maximum 24/7 stable fsb for this DFI board and what is the maximum bench stable?
Benchable 3D or 2D ??Quote:
Originally Posted by NoGuru
Just 2D so far. Still on air for a few more weeks.Quote:
Originally Posted by E-mil
That looks like a nice chip. Show us what you need to get 4.2-4.5GHz stable :)Quote:
Originally Posted by pvhk
Yeap - the RAM sucks, or I suck, and since I'm not willing to face the reality that it is me, it must be the RAM, I just can't squeeze any more performance out of it...on the other hand, I finally confirmed I'm 24hr Prime stable and incrementally lowerd my Voltages to 1.2875, thermals are much better - never above 69c and generally average 60-64c. (loaded) and idle around 39c
http://i209.photobucket.com/albums/b...sadler/ram.png
Code:GIGABYTE GA-EP45T_USB3P
[u]MB Intelligent Tweaker(M.I.T.)
Robust Graphics Booster ...............: AUTO
CPU Clock Ratio ..........................: 9X
Fine CPU Clock Ratio.....................:0.0
CPU Frequency ...........................: 4.05GHz GHz
Clock Chip Control
Standard Clock Control
CPU Host Clock Control..................: [Enabled] []
CPU Host Frequency (Mhz) ............:450Mhz
PCI Express Frequency (Mhz) .........:100Mhz
C.I.A. 2.............................:[Disabled]
Advanced Clock Control [Press Enter]
CPU Clock Drive...........................:800mv
PCI Express Clock Drive...................:900mv
CPU Clock Skew (ps).......................: 0ps
MCH Clock Skew (ps).......................: 0ps
DRAM Performance Control
Performance Enhance....................: [Standard]
Extreme Memory Profile (X.M.P.)........: DISABLED
(G)MCH Frequency Latch.................: [Auto]
System Memory Multiplier ..............:3.33D
Memory Frequency (Mhz) ..............: 1500Mhz
DRAM Timing Selectable ................: [Manual] []
Standard Timing Control
CAS Latency Time...............................: 8
tRCD ..........................................: 8
tRP...........................................: 8
tRAS...........................................: 24
Advanced Timing Control [Press Enter]
tRRD...........................................:AUTO
tWTR...........................................: AUTO
tWR............................................: AUTO
tRFC...........................................: AUTO
tRTP...........................................: AUTO
Command Rate (CMD) ............................: AUTO
Channel A
Channel A Timings Settings [Press Enter]
Static tRead Value......................:Auto
tRD Phase0 Adjustment...................:Auto
tRD Phase1 Adjustment...................:Auto
tRD Phase2 Adjustment .................:Auto
tRD Phase3 Adjustment..................:Auto
Trd2rd(Different Rank)..................:Auto
Twr2wr(Different Rank)..................:Auto
Twr2rd(Different Rank)..................:Auto
Trd2wr(Same/Diff Rank)..................:Auto
Dimm1 Clock Skew Control................:Auto ps
Dimm2 Clock Skew Control................:Auto ps
Channel A Driving Settings [Press Enter]
Driving Strength Profile:[AUTO]
Data Driving Pull-Up Level.............:Auto
Cmd Driving Pull-Up Level..............:Auto
Ctrl Driving Pull-Up Level.............:Auto
Clk Driving Pull-Up Level..............:Auto
Data Driving Pull-Down Level...........:Auto
Cmd Driving Pull-Down Level............:Auto
Ctrl Driving Pull-Down Level...........:Auto
Clk Driving Pull-Down Level............:Auto
Channel B
Channel B Timings Settings [Press Enter]
Static tRead Value......................:Auto
tRD Phase0 Adjustmen...................:Auto
tRD Phase1 Adjustment...................:Auto
tRD Phase2 Adjustment .................:Auto
tRD Phase3 Adjustment..................:Auto
Trd2rd(Different Rank)..................:Auto
Twr2wr(Different Rank)..................:Auto
Twr2rd(Different Rank)..................:Auto
Trd2wr(Same/Diff Rank)..................:Auto
Dimm1 Clock Skew Control................:Auto ps
Dimm2 Clock Skew Control................:Auto ps
Channel B Driving Settings [Press Enter]
Driving Strength Profile:[AUTO]
Data Driving Pull-Up Level.............:Auto
Cmd Driving Pull-Up Level..............:Auto
Ctrl Driving Pull-Up Level.............:Auto
Clk Driving Pull-Up Level..............:Auto
Data Driving Pull-Down Level...........:Auto
Cmd Driving Pull-Down Level............:Auto
Ctrl Driving Pull-Down Level...........:Auto
Clk Driving Pull-Down Level............:Auto
{U]Motherboard Voltage Control{/U]
CPU Vcore . ...........................:1.28750v
CPU Termination ...... 1.200V*...........: 1.2000v
CPU PLL .........1.500V*...........: 1.5000v
CPU Reference............................: .760
MCH/ICH
MCH Core ..........1.100V............:1.32
MCH Reference.. ... . .0.800V............; .800
ICH I/O .... ......1.550V............: AUTO
DRAM
DRAM Voltage ........1.500V............:1.34
DRAM Termination ... ..0.900V............: AUTO
Channel A Data Vref ...0.75V.............: AUTO
Channel B Data Vref ...0.750V............: AUTO
Channel A Address Vref 0.75V.............: AUTO
Channel B Address Vref 0.750V............: AUTO
[b]{U]Advanced Settings[b]{U]
Limit CPUID Max. to 3.....................: [Disabled]
No-Execute Memory Protect............: [Enabled]
CPU Enhanced Halt (C1E)................: [Disabled]
C2/C2E State Support....................: [Disabled]
x C4/C4E State Support..................: [Disabled]
CPU Thermal Monitor 2(TM2) ...........: [DISABLED]
CPU EIST Function.........................: [Disabled]
Virtualization Technology................: [Disabled]
Note: * refers to 45nm CPU
b1t, put 50 on your MCH skew, and raise your CPU termination, you will get a lot farther.
Thanks 'Guru, I'll check that out! - my understanding is that VTT should pretty much always stay behind the CPUv -is that right?Quote:
Originally Posted by NoGuru
Most of the time, I usually stay about .2 behind till you start getting really high clocks. So if my vcore is 1.40 then term will be about 1.38:up:Quote:
Originally Posted by b1tphr34k
yes true NoGuru i vtt should be closer to vcore :up:Quote:
Originally Posted by NoGuru
Hi! I just got my new Q9650 CPU ! I am using it on an asus maximus extreme x38 mobo! Right now it is OC to 3.6GHZ and cpu is quiet happy and cool!
here is my retail cpu information:
Pack date: 5/3/2010
FPO/Batch: L009B453
S-spec: SLB8W
VID: 1.2250v
http://img190.imageshack.us/img190/4658/64944802.png
http://img594.imageshack.us/img594/5329/sdc11149.jpg