Ebay is your best bet mate, that's how I got mine. I know you'll pay a premium, but the results are worth it ;)
Ebay is your best bet mate, that's how I got mine. I know you'll pay a premium, but the results are worth it ;)
Yes, greetings from Poland:)
Thanks In-Fluence for your suggestion. I'll look around although I'm quite afraid of buying a pig in a poke... I must admit that I used to avoid any internet auctions because of that. But will give it a try. Thanks for help ones again guys.
I have my E8500 @ 4Ghz, is it worth going to 4.3-4.4. my temps with my E8500 Lapped and my XIGMATEK Dark Knight Lapped are 30C-33c idle and at full load with Intel burn test 60C-65C. i use my pc for gaming not benchmarking
I bought my 45t-ud3r off ebay and am happy I did but it was never my intention to push it too hard I have just kept to the sort of voltages found in the templates above because there will be no rma
Well, it's indeed very nice to meet someone from as far afield as south america, who has polish roots. I bet you get tan as red as me and ur best meal is grilled homely wurst with relish and Tyskie beer heh:)Quote:
nice to meet a fellow from Poland, grandma and grandpa where from there, i live in south america but i dont look south american lol
As for beer, here's to you Guys. Cheers :toast:
So I've searched ebay a little and must say that not every seller has forgotten about UD3P yet :) In-Fluence, could u PM me? I'd like to ask u few questions concerning ebay. Thanks.
So I've been sifting through this thread for a few weeks now (MAN, it's a lot) and I'm still hung up. As my sig says it's a Xeon 3370 (Q9650) in a UD3R rev. 1.1 on the latest BIOS (just updated it 2 weeks ago).
Settings:
Vcore 1.4375 BIOS/1.41 OS, VID is 1.30
CPU term 1.28
PLL 1.57
CPU ref auto
MCH core tried everywhere from 1.28 to 1.38
MCH ref 0.75
Everything else is at auto except VDIMM which is at 2.0. Tried numerous straps keeping the RAM ~ 1000-1066 with 6-6-6-18 timings to make sure it wasn't getting in the way.
My question is, why can I not get it stable at 4.0? I started at 500FSB x 6 and Prime Blend passed 8 hours. 500x8 = spontaneous reboot, 446x9 = Prime errors. Temps are fine at the above volts, 29C idle/60C with 3 cores loaded. I know this rig has more in it; what am I forgetting?
Anyone here know of any other place's than newegg where you can get ahold of the ga-ep45t-usb3 mobo. Somewhere they ship to Europe?
Maybe that board isn't intended to hit the european market... most big etailers send worldwide, but is it worth the risk if something happens to it, that you have to send it all back to the States ?
Why not go for this one : EP45T UD3P
I know it's not intended for the European marked. And for people here that sucks. This is a newer mobo than the ep45-ud3p. Don't know if they have changed more than the max ram speed and usb3.
It's not worth sending back, no. If things don't work out the next step is something newer. But can anyone name some retailer somewhere maybe? I don't know of many.:(
Here's my first SPI 1M with my X3380 wish i have a kit of 2x1 of TeamX will be much lower for sure, also no tweaking at all pure hardware :D
http://img21.imageshack.us/img21/9168/sp11m12.png
Cheers :up:
Sergio
Amazing work!
Some AC to the rad...
i love this chip !!!
http://img717.imageshack.us/img717/4594/p10307411.jpg
http://img91.imageshack.us/img91/5416/aquamark37.png
Cheers
Sergio
Very nice work andressergio :up: A fine cpu indeed.
Has anyone been getting weird issues with their ud3p and prime95? I can run a full loop of large FFTs without issue, followed by another loop with small ffts, again completely stable. The moment I start a blend test, it fails within a minute or two of the first 1024k test?!
I'm a tad baffled now because it seems like I could have been stable all along and maybe prime has somehow corrupted or something weird with memory usage could be going on. Anyone else had similar issues?
ive noticed that when i use just small fft's i can prime for a couple of hours
and the same with large fft's.but when i do blend i get worker errors during
the transition between 1024k and 8k,or between 10k and 896k.
theres something about switching between small and large fft's that reveal
these error's that running just small or large doesnt.
when i error between 1024k and 8k usually a bump in vcore solves it.when
error between 10k and 896k a small increase in vdimm/vNB solves it.
Could someone point me to a guide that goes over the voltage settings for the UD3P? I'm not really understanding the reference voltages.
arfett,these limks should help you understand it.
lsdmeasap's mch reference guide
ET6 vs bios
http://i46.tinypic.com/2444b3c.jpg
Thank you for the links, but those are only about making Easy Tune show the right numbers. I'm looking for an explanation as to what those voltages are for and safe levels to set them, tips for stability, etc.
My temperatures stay in the 50s during LinX/Intel Burn Test and I definitely want to push for 4.5GHz, but I don't feel comfortable doing so without knowing I won't burn out my motherboard or ram with these reference voltages. I've combed through many pages in this thread, but must be missing this information if it's in this thread at all.
******Motherboard Voltage Control******
Load-Line Calibration: disabled (can be used if volt drop is to big at higher core speed. (I always use LLC when overclocked)
CPU Vcore: The voltage applied to CPU @ varying speeds in order to achieve stability
CPU Termination: 1.20 aka "FSB voltage", keeps your system stable at a certain bus speed, higher bus, higher CPU Term
CPU PLL: 1.50 (it helps you to boot into OS, higher bus speed, higher CPU PLL) 1.50 is good enough up to 450 FSB with a quad
CPU Reference: 0.760 (aka CPU GTL Ref, set this value when CPU Term is at 1,20 because it changes/scales with the CPU Term, this one is a bit tough to explain but in short, this one helps to have the CPU stable when increasing bus speed)
MCH Core: 1.20 -1.30 (aka NB voltage, this P45 setting is a bit sensitive, both too low or too high and it will not boot)
MCH Reference: (0.840?) (aka NB GTL Ref, set this value when CPU Term is at 1.20 because it changes/scales with the CPU Term, this one is a bit tough to explain but in short, this one helps to have the NB stable when increasing bus speed, this one is also a bit sensitive, one step up or down should be fine though)
MCH/DRAM Reference: Auto (reference between NB and ram, best set to auto) I use normal
ICH I/O: 1.50 (aka SB bus voltage, one or two steps up if you run raid or many HD)
ICH Core: 1.10 (aka SB core voltage, one or two steps up if you run raid or many HD)
DRAM Voltage: You know what to set for your ram
DRAM Termination: auto (termination voltage for the ram) I use normal
Channel A Ref: auto (reference voltage for ram slots A) I use normal
Channel B Ref: auto (reference voltage for ram slots A) I use normal
And nice job Serg!!!!!
Thanks dude. Are those numbers (.76 and .84) universally recommended for overclocking and could setting them higher cause stability issues or kill the motherboard?
Also here is what everything is set to after inputting those numbers with CPU termination at 1.2. Does this look safe to you?
http://img85.imageshack.us/img85/8528/et6.jpgCode:******Motherboard Intelligent Tweaker (M.I.T.)******
Robust Graphics Booster.........................[ Auto ]
CPU Clock Ratio....................................[ 8x ]
Fine CPU Clock Ratio..............................[ + 0.5 ]
CPU Frequency 4.037GHz...........................[ 475 x 8.5 ]
******Clock Chip Control******
CPU Host Clock Control..........................[ Enabled ]
CPU Host Frequency (Mhz)........................[ 475 ]
PCI Express Frequency (Mhz).....................[ 100 ]
C.I.A 2 [Disabled]
******Advanced Clock Control******
Default settings
******DRAM Performance Control******
Performance Enhance.............................[ Turbo ]
Extreme Memory Profile (XMP)....................[ Auto ]
(G) MCH Frequency Latch.........................[ Auto ]
System Memory Multiplier........................[ 2.00D ]
Memory Frequency ..............1066...... ...950
DRAM Timing Selectable..........................[ Manual ]
CAS Latency Time...............[ 5 ].........[ 5 ]
tRCD...........................[ 5 ]...............[ 5 ]
tRP............................[ 5 ]................[ 5 ]
tRAS...........................[ 15 ]..............[ 15 ]
******Advanced Timing Control******
Default settings
******Channel A Timing Settings******
Default settings
******Channel A Driving Settings******
Default settings
******Channel B Timing Settings******
Default settings
******Channel B Driving Settings******
Default settings
******Motherboard Voltage Control******
Load-Line Calibration...........................[ Enabled ]
CPU Vcore [ 1.28750 ]..............[ 1.325 ]
CPU Termination [ 1.200 ]...............[ 1.34 ]
CPU PLL [ 1.500 ]..................[ 1.5 ]
CPU Reference [ 0.760 ].................[ 0.848 ]
MCH Core [ 1.100 ]...................[ 1.3 ]
MCH Reference [ 0.760 ].................[ 0.848 ]
MCH/DRAM Refernce [ 0.900 ]..............[ 0.900 ]
ICH I/O [ 1.500 ].....................[ 1.5 ]
ICH Core [ 1.100 ]....................[ 1.1 ]
DRAM Voltage [ 1.800 ]................[ 2.1 ]
DRAM Termination [ 0.900 ]................[ 0.900 ]
Channel A Ref [ 0.900 ].................[ 0.900 ]
Channel B Ref [ 0.900 .....................[ 0.900 ]
*****************Advanced BIOS Features*********
Limit CPUID Max to 3 ..................[disable]
No-Execute memory Protect ........[disable]
CPU Enhance Halt (CIE) ..............[disable]
C2/C2E State Support ................[disable]
C4/C4E State Support ................[disable]
CPU Thermal Monitor...................[disable]
CPU EIST Function..................... [disable]
Virtualization Technology ............[enabled]
I know what you're getting at but that's not really the issue I'm having. Prime large & small are both ok, and OCCT is also the same after 6 hours. but the moment I start blend it fails in less than 5 minutes on random cores, no switching between 8 & 1024k, just as it starts out. Linx also went 5 passes fine so I would say the system could be classed as stable, but this is the one niggle I have and can only put it down to are the FEa bios, w7 64 thing or possibly the high 518x8.5 fsb. I'm now running the same tests at 521x8.5 with same settings and again, blend fails in the same amount of time but large FFTs seem to be 1024k stable. I'm gonna do a rebuild on the board soon, and hopefully the issue might dissappear but just wondered if others experienced the same thing? I'm guessing bios could have a lot to do with it.