1366 x58 Xeon Enthusiast overclocks club

primetime said:
First thing that comes to mind is 4.2ghz is a million times easier to get bullet proof stable over 4.5. It sucks but its the truth. Also from the way i understand it, running the ram at 2000mhz is hardly worth the extra trouble as well.....with that being said i would tackle one issue at a time if your convinced you need both
Click to expand...

I think that I'll use 4.4ghz in case if I woudn't be able to stabilize 4.5+ Seems (and feels) like 1.35v on core are enough for 4.4 w/o any tweaks and additional mb settings. But I'm still trying :D

Regarding memory, as I said before I fully tested it to be sure it's working. Hours of memtest, hours of prime & linx - no errors and problems at all. Also I tested it on desired multi with desired uncore, changing only core multi to down speed closer to default 3.3-3.4ghz to avoid any core overclock issues and errors in memory/uncore testing. Or you mean that memory at 2000 and core at 4ghz+ may require move Vcore bump to be stable? I'll check this also... I can set ram to 1600 and keep uncore higher than 2x (about 3500-3600) and see are there any difference in stability and Vcore required to hold it. But as I can see in this thread and other forums it's not so at all.

Seems my mem kit hate lower timings (CL8) when freq is above 1800mhz. I tryed to run 1800-8-8-8-24 and failed with 1.75v on mem. 1.65v is no boot. But it works fine on 9-10-9-27 timings, allows CR down to 1 instead of 2 default and tightening 2'nd timing significantly according to their default shown in XMP for 2000 mode. And all this with 1.6v instead of default 1.65v. I think I'll try to run 1.55-1.57 also, just for fun.

Tested 2100 mem freq this night. Seems to work... used bclock 150 and 1:7 multi for mem. 9-10-9-27 and 1.6v again.

Did some more research and realised that many people say that higher multies (any! - core, mem, uncore) - harder to stabilize the system and more volts are required. Seems that it's right and all this "free" multi are good for benching (maybe under dry ice or ln2) where you just changing multi and upping vcore. I'm not sure at all regarding memory multi (1:5 is better 1:7 for example and 1:6 is gold mid) but I should test it too. This way all this 30x+ multies seems unusable for 24/7.

This way seems that higher bclk still preferable and for 24/7 x27 multi (like max turbo for my W3680) is a bit high and I should move to 175-177 range. 1:6 multi for mem will give me 2100 that seems to work fine, while 4.4ghz is possible at 25x. Or move to "gold" s-1366 bclk at 200, lower mem multi to 1:5 for 2000 and have simple and more lower multies for core. 22-23.
 
I'll jump in on this!! I really love my X58 system. last real overclockers platform before skylake.
running a new R3E which is amazing. pulled the chip from a decommissioned server at a friends work.. just a lucky find. can't seem to get past 4.4 GHZ but i think i can live with it. I would really like to break 100 Gflops. Here's me current output.
12052375_443536739188109_1157128207857411997_o.jpg
 
Hey all.
Got my Xeon X5680 the other day. I would have done the 5670 but my P6T doesn't like Bclks past 185 or so, at least stable for gaming.

Using a Coolermaster Hyper 212 Evo with a 2nd fan added, Artic Silver ceramique(sp?) paste.

vCore: 1.3825v
CPU PLL: 188v
QPI/UCLK/Vtt: auto
vDDR: 1.66v
IOH/ICH: 1.25v
PCI-E: 1.5v

I am able to get 166bclk stable there. Temps hit mid/high 60's(low 70's for 2 cores) C full load prime or OCCT testing. Anything higher and it appears this chip wants quite a bit more voltage and I'm not sure I want to go that hard on this old motherboard.

Running 6x2gb Mushkin Redline PC1866 memory. I'm wondering if swapping to 3x8GB dims may help a bit vs the 6 dim load on the IMC???? I really don't want to spend much more on this rig though. For the $ it was a hell of an upgrade to make this box relevant again :).
 
be sure you are testing the max base clock the right way to start. drop your multi down to 13, memory below it's default speed, with the base clock figured in. so this way you are testing only the base clock. also remember that the xeons only require 1.5x uncore speed. choose the lowest available multi.

i would also pull out 3 of the memory chips and see if you gain any base clock head room. another thing to try is the clock skew adjustments. these have been known to allow some more stability at higher base clocks.
 
JNavy89GT said:
Hey all.
Got my Xeon X5680 the other day. I would have done the 5670 but my P6T doesn't like Bclks past 185 or so, at least stable for gaming.

Using a Coolermaster Hyper 212 Evo with a 2nd fan added, Artic Silver ceramique(sp?) paste.

vCore: 1.3825v
CPU PLL: 188v
QPI/UCLK/Vtt: auto
vDDR: 1.66v
IOH/ICH: 1.25v
PCI-E: 1.5v

I am able to get 166bclk stable there. Temps hit mid/high 60's(low 70's for 2 cores) C full load prime or OCCT testing. Anything higher and it appears this chip wants quite a bit more voltage and I'm not sure I want to go that hard on this old motherboard.

Running 6x2gb Mushkin Redline PC1866 memory. I'm wondering if swapping to 3x8GB dims may help a bit vs the 6 dim load on the IMC???? I really don't want to spend much more on this rig though. For the $ it was a hell of an upgrade to make this box relevant again :).
Click to expand...

That is Gobbs of Vcore voltage. i was able to get 4600mhz @ 1.379. you should be able to stabilize 4.4 @ 1.35 or so. that will bring the temps way down
 
I don't know how you guys hit 200mhz bclk but I can't get past 174 no matter what. I'm running a 5650 right now but I felt like just swapping for a 5670 and call it a day. I really want to be in the 4ghz+
 
Evga X58 SLI
Xeon x5650
Corsair H50
12gb Corsair dominator
Seasonic 750w
ATI 7870
128g SSD
 
~ever~ said:
I don't know how you guys hit 200mhz bclk but I can't get past 174 no matter what. I'm running a 5650 right now but I felt like just swapping for a 5670 and call it a day. I really want to be in the 4ghz+
Click to expand...

bclk all depends on the motherboard. I'm pretty sure the evga motherboards weren't known for their bclk clocks.
 
~ever~ said:
Evga X58 SLI
Xeon x5650
Corsair H50
12gb Corsair dominator
Seasonic 750w
ATI 7870
128g SSD
Click to expand...

motherboard is the factor here. one thing you could try if you are running 6 sticks of ram, pull half of them out for a test. But that X58 SLI is a pretty vanilla board.
 
Hi guys.

Having an issue getting past 166 bclk with my X5680 cpu.

cooling is Coolermaster Hyper 212 with 2 fans in push/pull

Temps at load are mid 50's to low 60's C which I find acceptable.

mobo is Asus P6T non deluxe with newest bios
cpu ratio auto
Speed step enable
Turbo-disable
Uclck freq=auto
QPI line-=auto
DRAM timing=auto
DRAM freq 1333(using 6x2GB Mushkin Redline PC1600 memory 8-8-8)
cpu 1.3625 vcore
qpi=auto
cpu pll 1.84
IOH 1.2
ICH 1.3
ICH PCI 1.5
Dram voltage 1.6(higher seems to cause issues)
IOH PCI 1.52

Loadline calibration-disable/auto is this ok? I have tried just as I'm typing and I'm going to see how that works. But I've read in various places LLC can kill the cpu. :/ . I have tried higher vcore w/out the LLC and stress tests(OCCT/Prime) still fail rather fast.

CPU and PCI spread spectrum disabled.
other's are auto if I didn't put it in here.

I think I'm out of the oc'ing scene enough that I'm overlooking something that's holding me back. I don't think it's the cpu as raising 1 MHz bclk will cause errors despite adjusting various voltages one at a time. vcore, vdimm, pll etc...
 
JNavy89GT said:
Hi guys.

Having an issue getting past 166 bclk with my X5680 cpu.

cooling is Coolermaster Hyper 212 with 2 fans in push/pull

Temps at load are mid 50's to low 60's C which I find acceptable.

mobo is Asus P6T non deluxe with newest bios
cpu ratio auto
Speed step enable
Turbo-disable
Uclck freq=auto
QPI line-=auto
DRAM timing=auto
DRAM freq 1333(using 6x2GB Mushkin Redline PC1600 memory 8-8-8)
cpu 1.3625 vcore
qpi=auto
cpu pll 1.84
IOH 1.2
ICH 1.3
ICH PCI 1.5
Dram voltage 1.6(higher seems to cause issues)
IOH PCI 1.52

Loadline calibration-disable/auto is this ok? I have tried just as I'm typing and I'm going to see how that works. But I've read in various places LLC can kill the cpu. :/ . I have tried higher vcore w/out the LLC and stress tests(OCCT/Prime) still fail rather fast.

CPU and PCI spread spectrum disabled.
other's are auto if I didn't put it in here.

I think I'm out of the oc'ing scene enough that I'm overlooking something that's holding me back. I don't think it's the cpu as raising 1 MHz bclk will cause errors despite adjusting various voltages one at a time. vcore, vdimm, pll etc...
Click to expand...

I've been runing both CPU, and QPI LLC since day one on my Xeon 5670. never had an issue. infact I read in a lot of places that the Xeon chiops love LLC, and can allow lower overall voltages. As far as the Base clock. That is almost completely dependent on the motherboard. First be sure you are checking for the true max base clock.

Turn down all multipliers to lowest setting, ram and uncore included, then try to boot with the base clock increase increments
 
Someone gave me two boards with bent pins, an Asus P6T and MSI X58 Platinum.

Thought I'd give fixing them a try.
Took a needle, a bright light, a magnifying glass, and went at it.

The P6T only had four bent pins and they were all bent to the side so it was pretty straightforward.
The MSI had five moderately (some up and down) bent pins as well as a pin bent backwards almost flat, I was able to get it back into place but I could feel that it was about to let go.

I dropped a w3520 in and fired them up.
The P6T runs flawlessly so far, passed Prime, etc, temps seem good. Still need to do more testing but pretty happy so far. Once everything checks out I'll be dropping a x5670 in there.
The MSI seems to run so far but I noticed the NB was running at 200 degrees (or so it said), I removed the heatsink and noticed the thermal paste was as hard as a rock, so I'm in the process of cleaning all that up. I'm not sure if I'll be changing the CPU on the MSI as I'm worried about that pin snapping.

All in all I probably have 45min in carefully pushing those pins into place (mostly due to the backwards one). :cool:

Edit: Cleaned up the thermal paste on the MSI and put some NT-H1 on there, now it's sitting at a much more reasonable 50c (122f). If I press on the heatsink a bit it drops to 45c, I don't think the springs are working quite like they should but 50c is acceptable without a fan, should drop quite a bit in a case.
 
Last edited:
I'm still trying to get stable 24/7 overclock, but seems something wrong... I can't undestand where the problem is and how to solve it.
I decided to stay at 4.4GHz max to keep VCore under or at 1.4v, also I decided to stay at max 2000 memory to avoid voltages over 1.65-1.66 on it.
Regarding Unclock and looking how it impacts performance but keeping in mind to not go over 1.35 QPI/DRAM I decided to left it in range about 3600-3800.
For mem on 2000 it will be lower than 2x multi, for mem like 1800 it will be 2x multi.

Last 3 days I thought I got my stable overclock but trying to start some tests again always resulted in errors/reboots.
I tried to stablilize 4.4GHz 200x22 1.39 vcore, with 2000-9-10-9-27-1T mem 1.65v, 3800 uncore 1.35v
CPU PLL I had tried was lowest possible 1.2v and default 1.81v

The problem is the following:
I run LinX with 100 iterations with 34000 problem size, it took 10 (!!) hours to complete and shown no errors.
Then I started Prime95 Tortute test with a custom settings (to allow more memory use in test, by default it starts with 2000mb of mem testing):
FFTs range 8 - 4096, mem to use 9216mb, time to test each FFT size 10 min (instead of 3min by default)
It run fine for several hours but then one of the worker threads stopped woth error (rounding 0.5 instead of 0.4)
It's always stop different worker and usually on different FFT size. Today it was 720k.

Could you give me an advice, where I should bump the voltage or maybe down a little?
When I give too low CPU Vcore with settings above, I may survive Linx as long as 5-20 iterations, but then it usually result in a blue screen or just reboots.
in 90% of time of low CPU vcore I get this message in windows logs:
Code: 
A fatal hardware error has occurred.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Internal Timer Error
Processor APIC ID: 4

The details view of this entry contains further information.

The computer has rebooted from a bugcheck.  The bugcheck was: 0x00000124 (0x0000000000000000, 0xffffe00014214038, 0x0000000000000000, 0x0000000000000000). A dump was saved in: C:\Windows\Minidump\102515-6109-01.dmp. Report Id: 102515-6109-01.

It's always give me 0x124 stop code and in 90% of cases with too low vcore voltage the processor APIC ID is 4.

Prime usually stop it's working threads with rounding errors - the more the lower voltage is and very rarely stop with blue screen or reboots. But in case of low QPI/DRAM voltage or too tight timings/high memory freq/uncore it's usually reboots. In most cases it logs the same 0x124 stop error, or logs nothing. But when it logs, the processor APIC ID is unrealistic - it may be 16, 18 etc (while I have only 12 of "CPU" id's)

When I stabilize Linx to pass 100 iterations in about 10-12 hours I may crash in Prime95 after 6-8-12 hours of tests. When I survived in Prime95 for 12+ hours I may crush in LinX... This makes me insane.

How long guys you tested you rigs usually?

Here is my current Rampage III Extreme BIOS template I work with (suggestions highly appreciated!!!):
Code: 
Target CPU Frequency		4400MHz
Target DRAM Frequency		2005MHz

LN2 Mode:			Disabled***
QPI Loadline Calibration:	Enabled***

AI Overclock Tuner		[Manual]
OC From CPU level Up		[Auto]

2D Performance Booster		[Disabled]
CPU Ratio Setting		[22.0]
CPU Turbo Power Limit		[Disabled]

>CPU Configuration----------------------------
CPU Ratio Setting		[22.0]
C1E Support			[Disabled]
Hardware Prefetcher		[Enabled]
Adjacent Cache Line Prefetcher	[Enabled]
MPS and ACPI MADT ordering	[Modern]
Intel(R) Virtualization Tech	[Enabled]
CPU TM Function			[Disabled]
Execute Disable Bit		[Enabled]
Intel(R) HT Technology		[Enabled]
*Active Processor Cores		[All]
A20M				[Disabled]
*Intel(R) SpeedStep(TM) Tech	[Disabled]
Intel(R) C-STATE Tech		[Disabled]
----------------------------------------------

BCLK Frequency			[200]
PCIE Frequency			[100]
DRAM Frequency			[2005]
UCLK Frequency			[3609]
QPI Link Data Rate		[7218]

Memory Recheck			[Enabled]
Memory Configuration Protect	[Disabled]

>DRAM Timing Control--------------------------
CAS# Latency		      9	[9]
RAS# to CAS# Delay	     10	[10]
RAS# PRE Time		      9	[9]
RAS# ACT Time		     27	[27]
RAS# to RAS# Delay	      7	[7]
REF Cycle Time		    160 [160]
WRITE Recovery Time	     14	[14]
READ to PRE Time	      8	[8]
FOUR ACT WIN Time	     28	[28]
Back-To-BackCAS# Delay	      0	[Auto]

Timing Mode		     1N	[1N]
Round Trip Latency on CHA    58	[58]
Round Trip Latency on CHB    60	[60]
Round Trip Latency on CHC    61	[61]

WRITE To READ Delay(DD)	      6	[Auto]
WRITE To READ Delay(DR)	      6	[Auto]
WRITE To READ Delay(SR)	     19	[Auto]
READ To WRITE Delay(DD)	     10	[Auto]
READ To WRITE Delay(DR)	     10	[Auto]
READ To WRITE Delay(SR)	     10	[Auto]
READ To READ Delay(DD)	      7	[Auto]
READ To READ Delay(DR)	      6	[Auto]
READ To READ Delay(SR)	      4	[Auto]
WRITE To WRITE Delay(DD)      7	[Auto]
WRITE To WRITE Delay(DR)      7	[Auto]
WRITE To WRITE Delay(SR)      4	[Auto]
----------------------------------------------

CPU Differential Amplitude	[Auto]
CPU Clock Skew			[Auto]
IOH Clock Skew			[Auto]
------------ Extreme Engine Digi+ ------------
Digi+ PWR Mode			[T-Balanced]
PWR Volt. Control		[10V]
Load-Line Calibration		[Full]
CPU Voltage OCP			[Enabled]
CPU PWM Frequency		[500KHz]

Extreme OV			[Disabled]
Extreme OC			[Auto]

CPU Voltage		1.382	[1.39375]
CPU PLL Voltage 	1.204	[1.20575]
QPI/DRAM Core Voltage	1.336	[1.34375]
DRAM Bus Voltage	1.640	[1.64300]

>DRAM REF Voltages----------------------------
DRAM DATA REF Voltage on CHA	[Auto]
DRAM CTRL REF Voltage on CHA	[Auto]
DRAM DATA REF Voltage on CHB	[Auto]
DRAM CTRL REF Voltage on CHB	[Auto]
DRAM DATA REF Voltage on CHC	[Auto]
DRAM CTRL REF Voltage on CHC	[Auto]
----------------------------------------------

IOH Voltage		1.244	[1.24550]
IOH PCIE Voltage	1.508	[1.51050]

ICH Voltage		1.138	[1.13950]
ICH PCIE Voltage		[1.51050]

---------- Spread Spectrum Control -----------
CPU Spread Spectrum		[Disabled]
DRAM Spread Spectrum		[Disabled]

My CPU-Z and CPU-Tweaker with settings above:
ODxn3IU.png
 
Last edited:
I want to put my 2nd X5650 to use in another build for an expansion on my home lab, but i hate how DDR3 is still expensive as i would want 48Gig :(
 
Gre4ka said:
I'm still trying to get stable 24/7 overclock, but seems something wrong... I can't undestand where the problem is and how to solve it.
I decided to stay at 4.4GHz max to keep VCore under or at 1.4v, also I decided to stay at max 2000 memory to avoid voltages over 1.65-1.66 on it.
Regarding Unclock and looking how it impacts performance but keeping in mind to not go over 1.35 QPI/DRAM I decided to left it in range about 3600-3800.
For mem on 2000 it will be lower than 2x multi, for mem like 1800 it will be 2x multi.

Last 3 days I thought I got my stable overclock but trying to start some tests again always resulted in errors/reboots.
I tried to stablilize 4.4GHz 200x22 1.39 vcore, with 2000-9-10-9-27-1T mem 1.65v, 3800 uncore 1.35v
CPU PLL I had tried was lowest possible 1.2v and default 1.81v

The problem is the following:
I run LinX with 100 iterations with 34000 problem size, it took 10 (!!) hours to complete and shown no errors.
Then I started Prime95 Tortute test with a custom settings (to allow more memory use in test, by default it starts with 2000mb of mem testing):
FFTs range 8 - 4096, mem to use 9216mb, time to test each FFT size 10 min (instead of 3min by default)
It run fine for several hours but then one of the worker threads stopped woth error (rounding 0.5 instead of 0.4)
It's always stop different worker and usually on different FFT size. Today it was 720k.

Could you give me an advice, where I should bump the voltage or maybe down a little?
When I give too low CPU Vcore with settings above, I may survive Linx as long as 5-20 iterations, but then it usually result in a blue screen or just reboots.
in 90% of time of low CPU vcore I get this message in windows logs:
Code: 
A fatal hardware error has occurred.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Internal Timer Error
Processor APIC ID: 4

The details view of this entry contains further information.

The computer has rebooted from a bugcheck.  The bugcheck was: 0x00000124 (0x0000000000000000, 0xffffe00014214038, 0x0000000000000000, 0x0000000000000000). A dump was saved in: C:\Windows\Minidump\102515-6109-01.dmp. Report Id: 102515-6109-01.

It's always give me 0x124 stop code and in 90% of cases with too low vcore voltage the processor APIC ID is 4.

Prime usually stop it's working threads with rounding errors - the more the lower voltage is and very rarely stop with blue screen or reboots. But in case of low QPI/DRAM voltage or too tight timings/high memory freq/uncore it's usually reboots. In most cases it logs the same 0x124 stop error, or logs nothing. But when it logs, the processor APIC ID is unrealistic - it may be 16, 18 etc (while I have only 12 of "CPU" id's)

When I stabilize Linx to pass 100 iterations in about 10-12 hours I may crash in Prime95 after 6-8-12 hours of tests. When I survived in Prime95 for 12+ hours I may crush in LinX... This makes me insane.

How long guys you tested you rigs usually?

Here is my current Rampage III Extreme BIOS template I work with (suggestions highly appreciated!!!):
Code: 
Target CPU Frequency		4400MHz
Target DRAM Frequency		2005MHz

LN2 Mode:			Disabled***
QPI Loadline Calibration:	Enabled***

AI Overclock Tuner		[Manual]
OC From CPU level Up		[Auto]

2D Performance Booster		[Disabled]
CPU Ratio Setting		[22.0]
CPU Turbo Power Limit		[Disabled]

>CPU Configuration----------------------------
CPU Ratio Setting		[22.0]
C1E Support			[Disabled]
Hardware Prefetcher		[Enabled]
Adjacent Cache Line Prefetcher	[Enabled]
MPS and ACPI MADT ordering	[Modern]
Intel(R) Virtualization Tech	[Enabled]
CPU TM Function			[Disabled]
Execute Disable Bit		[Enabled]
Intel(R) HT Technology		[Enabled]
*Active Processor Cores		[All]
A20M				[Disabled]
*Intel(R) SpeedStep(TM) Tech	[Disabled]
Intel(R) C-STATE Tech		[Disabled]
----------------------------------------------

BCLK Frequency			[200]
PCIE Frequency			[100]
DRAM Frequency			[2005]
UCLK Frequency			[3609]
QPI Link Data Rate		[7218]

Memory Recheck			[Enabled]
Memory Configuration Protect	[Disabled]

>DRAM Timing Control--------------------------
CAS# Latency		      9	[9]
RAS# to CAS# Delay	     10	[10]
RAS# PRE Time		      9	[9]
RAS# ACT Time		     27	[27]
RAS# to RAS# Delay	      7	[7]
REF Cycle Time		    160 [160]
WRITE Recovery Time	     14	[14]
READ to PRE Time	      8	[8]
FOUR ACT WIN Time	     28	[28]
Back-To-BackCAS# Delay	      0	[Auto]

Timing Mode		     1N	[1N]
Round Trip Latency on CHA    58	[58]
Round Trip Latency on CHB    60	[60]
Round Trip Latency on CHC    61	[61]

WRITE To READ Delay(DD)	      6	[Auto]
WRITE To READ Delay(DR)	      6	[Auto]
WRITE To READ Delay(SR)	     19	[Auto]
READ To WRITE Delay(DD)	     10	[Auto]
READ To WRITE Delay(DR)	     10	[Auto]
READ To WRITE Delay(SR)	     10	[Auto]
READ To READ Delay(DD)	      7	[Auto]
READ To READ Delay(DR)	      6	[Auto]
READ To READ Delay(SR)	      4	[Auto]
WRITE To WRITE Delay(DD)      7	[Auto]
WRITE To WRITE Delay(DR)      7	[Auto]
WRITE To WRITE Delay(SR)      4	[Auto]
----------------------------------------------

CPU Differential Amplitude	[Auto]
CPU Clock Skew			[Auto]
IOH Clock Skew			[Auto]
------------ Extreme Engine Digi+ ------------
Digi+ PWR Mode			[T-Balanced]
PWR Volt. Control		[10V]
Load-Line Calibration		[Full]
CPU Voltage OCP			[Enabled]
CPU PWM Frequency		[500KHz]

Extreme OV			[Disabled]
Extreme OC			[Auto]

CPU Voltage		1.382	[1.39375]
CPU PLL Voltage 	1.204	[1.20575]
QPI/DRAM Core Voltage	1.336	[1.34375]
DRAM Bus Voltage	1.640	[1.64300]

>DRAM REF Voltages----------------------------
DRAM DATA REF Voltage on CHA	[Auto]
DRAM CTRL REF Voltage on CHA	[Auto]
DRAM DATA REF Voltage on CHB	[Auto]
DRAM CTRL REF Voltage on CHB	[Auto]
DRAM DATA REF Voltage on CHC	[Auto]
DRAM CTRL REF Voltage on CHC	[Auto]
----------------------------------------------

IOH Voltage		1.244	[1.24550]
IOH PCIE Voltage	1.508	[1.51050]

ICH Voltage		1.138	[1.13950]
ICH PCIE Voltage		[1.51050]

---------- Spread Spectrum Control -----------
CPU Spread Spectrum		[Disabled]
DRAM Spread Spectrum		[Disabled]

My CPU-Z and CPU-Tweaker with settings above:
ODxn3IU.png
Click to expand...

To be honest, i think 99% of us settle for an easier overclock of 4.2 for year round use. Speaking for myself i was never 100% bullet proof at 4400mhz...very close..like 98% there but never 100% all the time. Something about that last 200mhz can only be seen in certain benchmarks anyhow;)
 
Honestly at 4Ghz+ you can't even feel a difference in use. No matter the game or program I run, I only see a difference in benchmarks after 4Ghz.
 
Okay, sorry for bumping this thread, but I've got a big issue at hands and it's getting really annoying...

I have a X5650 running on a X58 Sabertooth Asus motherboard, and whatever I do, it just refuses to do any voltage changes, I don't know what else to do.

I've disabled C-Stepping, turbo boost, IntelVT, hyperthreading... Nothing, it just does not accept it.

I will admit, I'm a noob in this platform, only dealt with 775 overclocking in the past, but there is something wrong here for sure as I've seen multiple people with my combo CPU+Mobo running this chip overclocked... I tried updating bioses, asked in bios mod forums, nothing. I'm getting really upset because I spent a good amount of money on this and am only able to squeeze out 3.4Ghz barely stable on the thing at max 1.16V (max turbo clock and voltage setting, BCLK OC).

Please, someone help me, I'm getting really upset over this, I'm even considering buying another processor just to see if that's what's causing this to happen...

My specs:

System specs:
Intel Xeon X5650 @3.41Ghz (6 Core/12 Thread) [Stepping 2 // Revision B1] SLBV3
Raijintek Triton 240 in Push
Asus Sabertooth X58
Kingston HyperX FURY 12GB DDR3 1600Mhz triple channel
EVGA GeForce GTX 680 SuperClocked 2048MB GDDR5 (1158Mhz)
Samsung 850 EVO 250GB SSD
Westen Digital Blue 1TB HDD
Western Digital Blue 500GB HDD
Corsair VS650 (I know it's shite)
Bitfenix Modcables

Bios settings:

7aYlyoI.jpg


(Yes, I have turbo enabled here but I tried disabling it along with other options in the bios, still nothing. And someone told me to turn down the QPI, I already did, it's running at 5k or 6k, something like that.)

AU5cLVX.jpg


(Applied 1.3V just to test out if anything changes..)

CynDLBz.jpg


(After rebooting into the bios, same old voltage, like nothing ever happend, again, tried disabling those features and still nada.)

qyDTY21.jpg


(Booted into windows, open CPU-Z to make sure it's not a motherboard misreading, nope, still there, stock as it could be. You can see it running at 3.1Ghz due to turbo being disabled, normally I have it running at 3.4Ghz)

If anyone can help me, It'd be gladly appreciated. :confused:
 
Something is wrong with your board, because it's not accepting new voltage value. Try bumping any other voltage like QPI or IOH, see if it sticks.
 
Vlad_13 said:
Something is wrong with your board, because it's not accepting new voltage value. Try bumping any other voltage like QPI or IOH, see if it sticks.
Click to expand...

Aw man, really? The only really "good looking" (for my build) X58 board I found x.x I will try and do that, but how can I check if they stay? Is there a monitoring tool for those voltages?

The thing is that, it's not even failing to boot after voltage changes, so it doesn't reset itself, it acts like it accepts the voltage changes and boots normally, but it doesn't change anything.
 
Mike Fox, try loading "setup defaults" in the exit tab -- once highlighted you have to press F5 then choose "Exit & Save Changes". Reboot, then change only the voltage, "Exit & Save Changes", then boot into windows and run something. Hopefully it'll be higher than 1.10. My P6T got into a funk where loading the setup defaults was the only thing that got it back.

Edit My main settings are for the x5670 with 48GB ram: BCLK of 180 for 3.96GHz, Turbo Off, Speedstepping on, C-states on. Vcore:1.28, QPI: 1.26, Vdimm:1.5
 
Last edited:
Does offset voltage work?

You might try removing the battery and letting it sit for 15min (while unplugged of course), I've had that resolve weird issues that simply resetting the bios didn't fix.

If none of that works, you might want to grab a cheap w3520 (i7 920) for $10 and see if that'll overvolt. If it does, it's probably the CPU. If not, then your board isn't working properly.
 
Ok, what about something really weird. Are you typing 1.300000 using the number keys or are you using the + and - keys? I'm wondering if it shows it being typed using the number keys, but in reality it needs the +- ones.
 
MN Scout said:
Mike Fox, try loading "setup defaults" in the exit tab -- once highlighted you have to press F5 then choose "Exit & Save Changes". Reboot, then change only the voltage, "Exit & Save Changes", then boot into windows and run something. Hopefully it'll be higher than 1.10. My P6T got into a funk where loading the setup defaults was the only thing that got it back.

Edit My main settings are for the x5670 with 48GB ram: BCLK of 180 for 3.96GHz, Turbo Off, Speedstepping on, C-states on. Vcore:1.28, QPI: 1.26, Vdimm:1.5
Click to expand...

I will be trying that out just now.

tbob22 said:
Does offset voltage work?

You might try removing the battery and letting it sit for 15min (while unplugged of course), I've had that resolve weird issues that simply resetting the bios didn't fix.

If none of that works, you might want to grab a cheap w3520 (i7 920) for $10 and see if that'll overvolt. If it does, it's probably the CPU. If not, then your board isn't working properly.
Click to expand...

None of them work, I am thinking of getting another processor to make things clear yes, but I'm yet seeing if I can get away without having to spend money.

MN Scout said:
Ok, what about something really weird. Are you typing 1.300000 using the number keys or are you using the + and - keys? I'm wondering if it shows it being typed using the number keys, but in reality it needs the +- ones.
Click to expand...

I am inputting the voltage with the numpad, not the +/- signs, tried both, none work.

Here's the video, I posted some screenshots of the actual video, but here it is:

https://www.youtube.com/watch?v=k-UHX7GQgB0
 
MN Scout said:
Mike Fox, try loading "setup defaults" in the exit tab -- once highlighted you have to press F5 then choose "Exit & Save Changes". Reboot, then change only the voltage, "Exit & Save Changes", then boot into windows and run something. Hopefully it'll be higher than 1.10. My P6T got into a funk where loading the setup defaults was the only thing that got it back.

Edit My main settings are for the x5670 with 48GB ram: BCLK of 180 for 3.96GHz, Turbo Off, Speedstepping on, C-states on. Vcore:1.28, QPI: 1.26, Vdimm:1.5
Click to expand...

Okay, update, tried doing that, nothing, max it hits is 1.2V (set it to 1.35) and with turbo, tried disabling everything on the bios, literally went mad, max it hits now is 1.096V.

Going back to my 3.1Ghz Config.
 
You certainly are in a unique situation.:( I'm not seeing anyone else in google search results with you same problem.

Do you have another PSU to test with? You are right on the ATX lower specification with your 3.3 at 3.1. My understanding is the 12v is used for the cpu circuitry, and there might be a chance 12V is nearly out of spec also.

The $10 ebay processor might be a better first test like tbob22 suggested.
 
MN Scout said:
You certainly are in a unique situation.:( I'm not seeing anyone else in google search results with you same problem.

Do you have another PSU to test with? You are right on the ATX lower specification with your 3.3 at 3.1. My understanding is the 12v is used for the cpu circuitry, and there might be a chance 12V is nearly out of spec also.

The $10 ebay processor might be a better first test like tbob22 suggested.
Click to expand...

Well even so, wouldn't the motherboard not care about the power supply? Like I understand that the power supply may not be enough, but wouldn't it accept the voltage changes regardless? As far as I know, there aren't motherboards who refuse to overclock because of a power supply not being enough, of course this being refusing to apply the settings, because not booting would be understandable, but it boots like nothing ever happened, it just ignores any voltage changes done to it.
 
have you tried doing a full clear c-mos? then load defaults save reboot, and try again? Just sounds like a bios glitch to me
 
Mike Fox said:
Well even so, wouldn't the motherboard not care about the power supply? Like I understand that the power supply may not be enough, but wouldn't it accept the voltage changes regardless? As far as I know, there aren't motherboards who refuse to overclock because of a power supply not being enough, of course this being refusing to apply the settings, because not booting would be understandable, but it boots like nothing ever happened, it just ignores any voltage changes done to it.
Click to expand...

I was thinking that maybe the motherboard power circuitry was stressing just to get it to 1.1V because of bad power.

With your 3.1GHz configuration. Does the voltage fluctuate? Change each time you reboot, or droop if you run a stress test?
 
MN Scout said:
I was thinking that maybe the motherboard power circuitry was stressing just to get it to 1.1V because of bad power.

With your 3.1GHz configuration. Does the voltage fluctuate? Change each time you reboot, or droop if you run a stress test?
Click to expand...

I don't think so. The voltage actually increases when it's under load to 1.16-1.18V..

If I disable the turbo boost and everything, it never goes above or under 1.096V.
 
primetime said:
have you tried doing a full clear c-mos? then load defaults save reboot, and try again? Just sounds like a bios glitch to me
Click to expand...

How would I do that? Remove the battery and stuff?

I've tried that but I'll give it another try today, leave it unplugged without the battery for like 15 minutes or so and see if that helps.
 
Mike Fox said:
How would I do that? Remove the battery and stuff?

I've tried that but I'll give it another try today, leave it unplugged without the battery for like 15 minutes or so and see if that helps.
Click to expand...

yea this is what i was thinking.....i also hold the power button a couple times...but yea remove power, battery, switch jumper for a little bit. Are you using the latest bios? Maybe a different bios version might help.
 
primetime said:
yea this is what i was thinking.....i also hold the power button a couple times...but yea remove power, battery, switch jumper for a little bit. Are you using the latest bios? Maybe a different bios version might help.
Click to expand...

I will probably be jumping to 1302 version of the bios, I have the 1402 version.
 
Has this issue been resolved to where Windows 10 does not show overclock speed. Please see below? My Windows 10 partition does not show proper speed, but my MAC OSX10.11.1 does. I tried downloading ChrisPC Win Experience, but that did not help..any advice?

Unable to post pics due to limited time being on forum...sorry

http://imgur.com/aE9NLPq


http://imgur.com/4mx9LzQ

Can someone please take a look at my overclock settings. It seemed to be pretty stable until today where CSGO crashed on me. I have since lowered my Bclock to 201 instead of 211 giving me a speed of 4.2ghz. I got these settings from doing research. my CPU temp is at about 80-85c at full load. Let me know How I did and what you would do to improve. I'm on a water-cooled system. I'm concerned about my PCIE setting...more than anything. This is my first overclock so please bare with me.

http://imgur.com/OPzb4hK

http://imgur.com/XikXaID

http://imgur.com/CuCHF9L

http://imgur.com/z4HkZeF

http://imgur.com/BFxgfBU

http://imgur.com/qHxungs

http://imgur.com/7uT42cy

http://imgur.com/aeWdx1V
 
Last edited:
Last edited:
As an eBay Associate, HardForum may earn from qualifying purchases.
You must log in or register to reply here.
Back
Top