BSOD help driver_IRQL

vietkangta

Limp Gawd
Joined
Dec 18, 2005
Messages
316
Hello, im running x5690 and having this bsod occur randomly. I've ran multiple different oc stress test and they all pass fine. memtest was ran for 9 hours.

i read error code ending in d1 has to do with dram/qpi voltage and ive been upping that. was at 1.33 and it bsod so now im at 1.35

I am trying to run my memory at the rate speed of 1600. My temps are all good.

Does anyone know what i need to do. My cpu is at 4 ghz at 1.3 vcore with bclk of 160. So not really pushing that much


CPU Ratio: 25
BLCK: 160
PCIE: 100
DRAm Freq: 1604
UCLK: 3207 (double)
QPI: 5774 (set to lowest)
CPu Volt: 1.3
Cpu PLL: 180
QPI: 1.33 (1.35 now since it just crashed with 1.33)
RAM Voltage: 1.66

Everything else on auto.

Ram timings ran as suggested by CPUZ for 1600

==================================================
Dump File : 062920-34015-01.dmp
Crash Time : 6/29/2020 10:46:04 PM
Bug Check String : DRIVER_IRQL_NOT_LESS_OR_EQUAL
Bug Check Code : 0x000000d1
Parameter 1 : fffff805`29fc43a4
Parameter 2 : 00000000`000000ff
Parameter 3 : 00000000`00000000
Parameter 4 : fffff805`29fc43a4
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+1c23a0
File Description : NT Kernel & System
Product Name : Microsoft® Windows® Operating System
Company : Microsoft Corporation
File Version : 10.0.18362.900 (WinBuild.160101.0800)
Processor : x64
Crash Address : ntoskrnl.exe+1c23a0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\WINDOWS\Minidump\062920-34015-01.dmp
Processors Count : 12
Major Version : 15
Minor Version : 18362
Dump File Size : 780,878
Dump File Time : 6/29/2020 10:47:37 PM
==================================================
 
Last edited:
That's a heck of an overclock for that CPU. Have you tried backing it off a bit. I've found that particular error to often be caused by a faulty overclock.
 
Agree with everyone else that you need to dial your CPU overclock back.
 
5Ghz is a lot for those old 1st Gen Core CPUs. How long do you run your OC stability tests for? Your OC might be "mostly" stable.

Sorry it was suppose to say 4ghz. i run a bit of everything For several hours . With prime95 i ran about 8 hours
 
Last edited:
That's a heck of an overclock for that CPU. Have you tried backing it off a bit. I've found that particular error to often be caused by a faulty overclock.

really? I did a a lot of reading before i set it to 4ghz most people with the chip did 4ghz easily with lower voltages than me.
I was under the impression that it should be easy to hit 4ghz
 
really? I did a a lot of reading before i set it to 5ghz most people with the chip did 4ghz easily with lower voltages than me.
I was under the impression that it should be easy to hit 4ghz

You should tone it down and see if it BSODs on your stress testing, you will then know what the problem is. Generally when troubleshooting you want to remove all OCs and see if it's stable first.
 
what RAM kit are you using?. .1.35V? unlikely for DDR3 RAM.. those tend to use 1.5V - 1.65V between 1600mhz and 2400mhz... 1600mhz are typically 1.5V

QPI/VTT depend mostly on what method of overclocking are you using? what BCLK and multiplier are you set? you may be required anything from 1.35V to 1.40V there if you are going high on clocks.
 
what RAM kit are you using?. .1.35V? unlikely for DDR3 RAM.. those tend to use 1.5V - 1.65V between 1600mhz and 2400mhz... 1600mhz are typically 1.5V

QPI/VTT depend mostly on what method of overclocking are you using? what BCLK and multiplier are you set? you may be required anything from 1.35V to 1.40V there if you are going high on clocks.

My ram are running at 1.66 even though the kit says 1.5v

CPU Ratio: 25
BLCK: 160
PCIE: 100
DRAm Freq: 1604
UCLK: 3207 (double)
QPI: 5774 (set to lowest)
CPu Volt: 1.3
Cpu PLL: 180
QPI: 1.33 (1.35 now since it just crashed with 1.33)
RAM Voltage: 1.66

Everything else on auto.

Ram timings ran as suggested by CPUZ for 1600
 
then like kirby said, back to stock and make sure its stable. then work your way back up.
 
My ram are running at 1.66 even though the kit says 1.5v

CPU Ratio: 25
BLCK: 160
PCIE: 100
DRAm Freq: 1604
UCLK: 3207 (double)
QPI: 5774 (set to lowest)
CPu Volt: 1.3
Cpu PLL: 180
QPI: 1.33 (1.35 now since it just crashed with 1.33)
RAM Voltage: 1.66

Everything else on auto.

Ram timings ran as suggested by CPUZ for 1600

IIRC 133x26 it's the default CPU multiplier, so you are overclocking basically by the BCLK 160*25 = 4000, you can decrease BCLK and increase multiplier and achieve the same frequency (if you don't want to keep increase QPI Voltage) 144mhz BCLK * 28 will achieve the same clock and probably will also be more stsable.
 
Your config is totally wrong:

* QPI data strap also matters and should be higher than uncore but less than 8 MT/s.
* For Westmere, it is better to push DDR clock high as possible (increase timing if needed). Uncore can be in 1.5x-2x ratio.
* QPI/VTT voltage should be maximum at 1.3v for 24/7 use.
* Vcore 1.325v (with LLC) for 24/7.
* DDR voltage 1.65v for 24/7. Samsung DDR3 chips are the best. Even the cheapest one can hit 2000 MHz.
* Use HCI Memtest (one instance per thread) or MemTest64 or Linpack Xtreme to test memory stability.
* Use Prime95 v26.6 custom FFT run of 512k to 1024k with 90 percent of RAM to test uncore stability.
* Use Linpack Xtreme to stress test the processor cores.

I suggest at least 8 hours for each component. Start with DDR and uncore.
 
also depending on how long the overclock's been set there's a chance that there's some degradation on the power delivery for the motherboard. if none of the other suggestions help that could be a possibility.
 
X58 motherboards are resilient... almost like WD 80GB IDE HDDs. Uncore voltage is the killer of the platform.

There are plenty of BAD guides on the Internet claiming "1.35v is fine! DDR voltage doesn't matter as long as its in the 0.05v from QPI/VTT".
 
Your config is totally wrong:

* QPI data strap also matters and should be higher than uncore but less than 8 MT/s.
* For Westmere, it is better to push DDR clock high as possible (increase timing if needed). Uncore can be in 1.5x-2x ratio.
* QPI/VTT voltage should be maximum at 1.3v for 24/7 use.
* Vcore 1.325v (with LLC) for 24/7.
* DDR voltage 1.65v for 24/7. Samsung DDR3 chips are the best. Even the cheapest one can hit 2000 MHz.
* Use HCI Memtest (one instance per thread) or MemTest64 or Linpack Xtreme to test memory stability.
* Use Prime95 v26.6 custom FFT run of 512k to 1024k with 90 percent of RAM to test uncore stability.
* Use Linpack Xtreme to stress test the processor cores.

I suggest at least 8 hours for each component. Start with DDR and uncore.

Hi to address your points

- my qpi is less than 8000mt/s. It’s around 5000
- i am trying to run my ram at 1600mhz which is the rated speed of the ram. And I’m running the uncore at 2x as recommenced by people with same chip and motherboard
- qpi volt is at 1.35. It appears to be more stable as i increased it from 1.30. But i can dial it back down and change other settings
-v core at 1.3 as it didn’t make a difference when i increased this
- ddr voltage is 1.64
- I’ve used multiple stress test and none of it gives error yet or crashes. The times I’ve crashed were either light load gaming. Or my pc Is idling on Desktop with no activity, this has happens twice.

Blender - 2.5 hours
Prime95 blend - 8 hours
Prime95 small ftt - 8 hours
Hci memtest - 12 hours
Memtest64 - 9 hours
Aida64 stress test - 7 hours
Intel burn test max - 10 passes
Linpack Not extreme - 2 hours
Realbench - 2 hours
and lots of gaming benchmark tests.
 
Back
Top