Testing later tonight.
Gotta go look at a truck (might buy it) and maybe re-lap my TRUE.
3.5 @ 1.264v.
440fsb works perfect.
450 crashes any settings.
P5K-DLX Wifi/AP 0603.
Printable View
Testing later tonight.
Gotta go look at a truck (might buy it) and maybe re-lap my TRUE.
3.5 @ 1.264v.
440fsb works perfect.
450 crashes any settings.
P5K-DLX Wifi/AP 0603.
449 = good
450 = bad.
I think its pretty obvious its bios related now thats holding most back from going 450+
That is why I sold my Maximus for a Rampage. My Maximus could only do 445 FSB. Rampage can do 480-485 FSB with a X3350.
But Intel last week announced that they are going to a E0 Core Stepping on all Yorkfield quads in the near future.
I've hit about the exact same wall on my 3350 as stevil. I'm running 448 right now, but going to 450 will pass 3+ hours of prime but fail overnight folding. I've increased ALL voltages available 2 steps higher to try to get 450 stable and I still get the same results :( Might have to invest in a P45 board if they clock better.
Kensek, I have just purchased ASUS Rampage Formula Motherboard and a TRUE with a 4,000 RPM fan on it, I can't wait for my X3350 to arrive! Please can you tell me what settings you used to get it to 3.8GHz, if you did? Many thanks, TehPwn!
http://img99.imageshack.us/img99/234...50at380qz7.png
Here are the 480FSB settings on my Rampage.
I also skewed the memory
Channel A +100
Channel B +200
Add a little to the VCore due to VDroop about .03v
I thought I fixed the problem by manually setting my ram's timing, fsb, and voltage because it survived an hour of priming(this was like two days ago). Today I attempt to lower the voltage, but it seems like none could survive within 5 minute of priming. So I went back to 1.45 volt on bios and I thought it still works. Just to make sure, I decide to prime it again for an hour before I prime it for the whole night when I sleep tonight. For some reason, when I went back to 1.45, my pc got a BSOD after 1 minute of priming. I can't think of anymore solution for this. It would really be nice if someone can lead me to enlightenment for this overclock. Also, can someone recommend me a setting I could go with?
It would be nice if someone can provide me a program that can tell me what volts I'm at, not just cpu volt.
P.S. For those who haven't read my case, just go to the page before this and you'll find it.
I haven't been following this thread but dropped in because I'm running the X3350. Here's what I've go it upto on the ASUS 790I. Note this is unlinked as my RAM is not stable past 1610mhz
Kensek, thanks. So its prime stable at these settings? Do you think it could do more? Just out of curiosity, what power supply are you using? Thanks.
I use a Corsair 750TX (750 Watter) in this rig. It is a single 12v Rail PSU, which is much better than those split 3-5 12v Rail PSUs.
I've never run Prime. I just run benchmarks at the outermost limit and then just run the CPU at 3.40 GHz. No need to run a CPU 24/7 with so much stress on it.
Tomorrow the Foxconn BlackOps Mobo comes along with an E8400. The DDR3-1866 Ram came today. And the TRUE came yesterday. Will play with the E8400 over the weekend.
Would have loved to try the X3350 on it but I sold all of mine. Wanted a Quad that can do 4.0 GHz on Air. So far I've had no luck with a few X3350s and a QX9650.
My QX9650 could only do 3.95 GHz so I sold that too 3 months ago.
Wow! The BlackOpps sure is nice! But quite expensive LOL! What do you think was limiting your X3350? Its VID, or the cooling? I think you may get 4GHz on it with LN2 or Dry Ice maybe, but remeber the 12MB Quads are 400MHz faster clock per clock than the 8MB ones, so a X3350 at 3.8 would be like a Q6600 at 4.2GHz.
I just use Air Cooling with a Zalman CNPS9700NT so I do not really push the CPU voltage. Hence I do not know what is holding back a X3350 from running at 500x8. I think it is the chip itself, and that is why Intel is going to a R0 Core stepping. Plus some of it is the Mobos BIOS. Like I said my Old Maximus could only do 445 FSB. My Old P5E3 Premium could only do 460, hence I sold it after 4 hours of use. But the Rampage is the best at reaching the highest X3350 FSB. But since they only peaked at 3.84GHz I sold them on eBay for $425.
Might try a QX9650 again with R0 Core stepping when they become available in 2-3 months.
The Lapped TRUE that I bought from CrazyPC.com will be my first CPU HSF that is not a Zalman. I have 2 9700's and 1 9500 that I use in other builds. But everyone claims these Thermalright Ultra 120 Extremes are the best and the reviews seem to back up those statements.
Whats so gay is, why on earth do fliiping Intel start their lowest range Yorkfield with a 8 Multi opposed to the lowest Kentsfiel which has a 9x Multi!! Damn, even the Q9550 only has an etra .5x Multi, and thats like neary twice the price!! Stupid!! LOL! Im really suprised that a QX9650 couldn't get 4GHz as ive seen people on Youtube with 790i boards (worse for overclocking Quads than X48's) gett 4.2-4.4GHz! Weird!! But good luck with your testing!!, TehPwn!
Due to their higher FSB they can't start with very high multi'sQuote:
Originally Posted by TehPwn
This to limit max overclock speed ofcourse... if they gave everybody 10 multi everyone in here would run at least 4ghz... and some with excessive voltage to get there... overall we get about same CPU speeds, but with bigger cache, new set of instructions, requiring less volts and a cooler running CPU...Quote:
Originally Posted by TehPwn
And they sell like hot pancakes... better await Q9645 then as it will be 9 multi then ( in fact a locked QX9650), this possibly should retail at around 550dollars...and drop down prices on lower specced CPU's and 4ghz will be in reach...
Guys , is there a date for the R0 release ? Im looking for a new toy to play with :up:
Maybe , a friend of mine has his highest ever q6600 clocks on a 780i so I expect a 790i will probably be even better.Quote:
Originally Posted by TehPwn
Jeez I should have kept my x3350 then , I might have to buy another(or 3360/q9650)Quote:
Originally Posted by TehPwn
:up:
i really doubt that x3350 @ 3.8ghz = Q6600 @ 4.2ghz.. at least for 3Dmark06.. in my case, with same hardwares (gfx, ram, mobo, psu, os..) my X3350 needs 3.76Ghz to finally exceed my Q6600 score @ 4GHz (14091 vs 14075) .. or in terms of cpu score 6058 vs 6017
at 3.72GHz, the x3350 only score 14059 in 3Dmark06 (cpu score: 5953)..
so i guess, x3350 needs to be near to 4GHz to equal the power of Q6600 @ 4.2GHz..
LOL! I just read that someone had Super Pi on both a Q9450 and a Q6600 in Super Pi needed a clock speed of 3.8GHz to get equal to a Q9450 at 3.4GHz, but i dunno about 3DMark, although I do use alot of 3D Mark :(
Yea, anyone got a release date for R0 stepping or Q9645? Thanks.
P.S Kensek did you have LLC enabled or disabled? Thanks. So from what we've learnt a X3350 at 3.8 is like a Q6x00 @ 4.05GHz, or around there, are people getting above 4GHz on the Q6700's? Cause if I could get 4.1-4.2 on a Q6700 which has a 10x Multi for £70 less, then it may be worth me getting a Q6700 instead??? Thanks.
I keep LLC (LoadLine Calibration) to whatever the default BIOS uses.
Wow 4.1-4.2 on a Q6700. I bought a Retail Q6700 last month when the price dropped to $266 and sold my Q6600 which could do 433x9. I thought your way - that a 10 multi would get you to 4.0Ghz.
Wrong, at least for me. I tried it in an Asus Rampage and a P5E3 Premium. Best it could do was 385 x 10 with the CPUv at 1.75v. Was afraid to go higher.
The Q6600 did 433 x 9 at 1.525v.
Today I spent 8 hours building the FoxConn Blackops. This board to me is a Major Disappointment.
The Board layout sucks.
1) Ram Release tabs cannot be fully pressed down when a graphics card is in PCIe slot1.
2) Their cheesy Sound Card blocks the access to one of the Mobo hold down screws and inhibits access to removing the CR2032 Battery.
3) The Manual does not tell you what any of the digital error codes mean.
4) I'm getting error code "E0" and the board will not POST or Display anything.
It may be the Patriot DDR3-1866 Ram. Zipzoomfly shipped it in a Tyvek Fedex Bag (Non bubble wrap protected) and it is packaged only in cheap formed hard plastic.
Quote:
Originally Posted by TehPwn
i used Q6700 and Q6600 before and both has same clock limit around 4.1GHz on water. I bought Q6700 because of the extra multiplier but i still had to lower down the multiplier to 9x to achieve higher clock. With 10x multiplier I only managed to hit around 380fsb..
It's June and still no stinkin' fix from :banana::banana::banana::banana:in ASUS. :mad: If they don't fix it this month I will NEVER buy anything again with ASUS written on it, I swear.
Who would you buy from them?
My Foxconn Blackops Mobo sucks - poor board layout. May be it will have superior operational abilities, once the thing gets beyond the E0 error the Digital Bios Debug LED Indicators are showing. Which Foxconn told me is a Memory error. Hence, I bought some cheap DDR3-1066 2x1GB sticks from Newegg for $70 to see if they will bootup. The Patriot Viper DDR3-1866 will not.
Hopefully it is just increasing the voltages in BIOS when I get into it with the cheap DDR3 so that I can use the 1866 memory.
Abit - No Support - 10 times worse than Asus.
Gigabyte - Bios revision every 2-3 months. 2-3 worse than Asus.
Asus still has 55% of the Global Motherboard Market Share.
Because they supply OEM's like Dell (MSI does as well) not because they make good boards (they dont).
Just realized I had a Q9450 on preorder somewhere and it's processing now - should I give it a try? Can I expect anything different from this later shipment (my X3350 is an early stuff, from ~mid-March.)
No.Quote:
Originally Posted by T2k
Hmmm. Not a chance, right?
I'll receive my X3350 along with a Asus P5E this Friday. It will run on air under a CNPS9500 AT until I get my water cooling setup completed.
Good luck. Let us know if you manage to break the 440 FSB wall.Quote:
Originally Posted by AceGoober
System
P5E3 Deluxe WiFi Bios 1109
X3350
OCZ PC 12800
Corsair Airflow
Ultra Extreme 120
Panaflo 120x38 103.8 CFM
eVGA 8800GTS KO 512MB
Zalman 750HP
Adda 120x25 88 CFM
----------------------
VCore 1.30V
FSB Strap AUTO
FSB 470
PCIE 100
DRAM Comand Rate 1N - 7-7-7-20
CPU PLL Voltage 1.58V
FSB Termination Voltage 1.34V
DRAM Voltage 1.98V
NB Voltage 1.57V
SB Voltage 1.20V
Clock Over-Charging Voltage: 0.80
Load Line Calibreation: ENABLED
CPU GTL Volt 0.67x
NB GTL Volt 0.67x
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
C1E Support: Disabled
CPU TM Function: Enabled
Execute Disable Bit: Enabled
Max CPUID Value: Disabled
Vanderpool Disable Bit: Enabled
http://img136.imageshack.us/img136/5...asprimeqb2.jpg
475 FSB
VCore 1.33125V
FSB Strap AUTO
PCIE 100
DRAM Comand Rate 1N - 7-7-7-21
Ai Clock Twister: Auto
Ai Transaction Boster: Enabled
Boost Level [2]
CPU PLL Voltage 1.60V
FSB Termination Voltage 1.48V
DRAM Voltage 2.06V (Corsair Airflow )
NB Voltage 1.67V
SB Voltage 1.20V
Clock Over-Charging Voltage: 0.80
Load Line Calibreation: ENABLED
CPU GTL Volt 0.67x
NB GTL Volt 0.67x
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
C1E Support: Disabled
CPU TM Function: Enabled
Execute Disable Bit: Enabled
Max CPUID Value: Disabled
Vanderpool Disable Bit: Enabled
http://img136.imageshack.us/img136/9...prime95cb7.jpg
Although all these results, unhappyly I do not obtain to twirl no version of 3Dmark. :down:
I moved for BIOS 1203 and was bad, is not ready still.:shocked:
Sorry for my bad english.
Ive just put my X3350 in my system today, and havent tried Overclocking yet but I have a weird situation. Core Temp is reporting my VID as 1.25v!!!! BIOS says default voltage is 1.200!!! LOL, I hope this is wrong, I thought Xeon's always have lower VID's!!!????
I wouldn't care about the VID too much. People with high VIDs get good overclocks sometimes while ppl with low ones don't. If you are unlucky enough to have an X38 or x48 board you have a good chance that the board will limit way before you reach the CPU's limit as well.Quote:
Originally Posted by TehPwn
Is the wall with the motherboard or the processor?
Quote:
Originally Posted by lowtcell
Pretty sure it's the motherboard.Quote:
Originally Posted by AceGoober
OK, it's up to 3.36Ghz, 417 @ 1.22v and it's rock solid stable. I'll bump up the FSB slightly until I hit a theoretical wall then post the results.
Quote:
Originally Posted by lowtcell
Is there still no fix from Asus for the P5E/Rampage for the 450 FSB wall? :mad:
No.Quote:
Originally Posted by FiXeR
I think I am going to have to change motherboards. This P5E won't go over 417 FSB no matter what I do. :(
holy crap in a basket... AceGoober where the hell have you been? i haven't seen you since DFI-Street was still running.
ya know you could sign into Yahoo once in a while.
you still running that big arse plasma on your main rig?
Quote:
Originally Posted by FiXeR
? Plenty of Rampage boards do +450Mhz FSB on quad-cores without any problems. No need for a fix to get over 450Mhz FSB. They might need some tweaking and more NB voltage, but if there is anything at 450Mhz FSB it's more like a simple fence instead of a wall. :)Quote:
Originally Posted by mu||et
I second that. Broke 450Mhz long ago.Quote:
Originally Posted by Abducted
I can run my X3350 easily at 3.4GHz (425x8) with stock voltages, though anything past that becomes a major issue. I had it briefly at 440x8 though with voltages that I surely wouldn't want to run 24/7 and it still wasn't stable.
After reading all the threads about the Maximus Formula and X38 chipset having a FSB wall at 440MHz, I'd have to say its spot on for me. I tried to run 450FSB and things just got all weird. It would only POST sometimes and other times it'd freeze in the BIOS menus.
Sadly I doubt ASUS really cares since the Maximus Formula seems to be discontinued.
Finally figured out the issues with overclocking and it is now running happily along at 3.45Ghz. Even though at stock voltage for this X3350 is 1.1 volts I had to let the motherboard control the voltages instead of doing it manually. I've got it running with Prime on small FFTs with OCCT going as well. Will post results when it gets done with a 12 hour run.