@eivrah

The one that binged your board back is already modified.
It doesn't have these changes on the QPI configuration but it should let you select higher TRFC ( 200 instead of the default 140 ) for memory like 1600+
It also have NVME support, updated microcode (1F) and higher IRST version (15.0.0.2371), so, if you plan to use the RAID function use the software/driver version 16.0.1.1018 should work without issue, or if it does change the "MessageNumberLimit" to 8.
The one I'm testing work in the same way as having the QPI setting in the memory because there something else overriding the user input value once a certain value/item is selected.
On your board the original user menu already work as it should, or at least it seems, since you can change the host clock frequency.

Mine just doesn't boot past 136 with anything tried so far.
Here the "stock" ( @135 default bios setting ) X5675 on this board.

@noInk I also notice the RAM settings is not sticking whenever I put custom values it doesn't boot my multiplier was always set to Auto.

@eivrah

Yep, but it's happening on every bios version I had the time to test, so far 5020 > 5600 ( there no support for more then 8 processor on anything before )

I think mine issue is introduced in 4104 and exacerbated with the changes first introduced on 4196 ( might check with the i920 down the line )

4104 Fixed issue where Intel® Desktop Control Center fails to launch after HCF value is applied to 136MHz or above.
4196 Added XMP support using BCLK override.
 

Your and also everyone else issue start just on 5020 Gulftown release (Memory Multiplier must be set on Auto)

5020 When selecting Automatic mode the Uncore Multiplier was being set to 22 instead of the proper system default.
5020 When selecting Manual mode the minimum supported ratio between Memory Ratio and the Uncore Multiplier will be used based on the processor installed.
 

There you have it, last working bios release doesn't exist. They didn't fix anything since 2127 ( the less broken release validated on real hardware ).
Intel just introduced broken workaround and new feature - later made working only on "auto".

CPU-Z
Yes too bad my RAM is my limit this cpu could have OC 4.5Ghz

@eivrah

Use this ratio to keep the QPI around 3200 till 166MHZ base clock:

"Uncore Multiplier" to 20
"Memory Multipler" to "auto" (changing this to any other value will likely make the board unbootable)
"QPI Data Rate" to 5.8

From your CPUz page I see the North Bridge @ rev 13, mine is rev10 but shouldn't be an issue with the overclockability.

I didn't seem quit understand the QPI's yet. I think I can't pass above 150 multiplier, my memory might be my limiter. I'm on 150 Right now + Turbo Boost enabled. Running on a Hackintosh so far stable.

@eivrah

The QPI Multiplier or "QPI Data Rate" on this board work as follow:

x20 = 4.8 GT/s
x22 = 5.866 GT/s
x24 = 6.4 GT/s

Who named those on the user interface just didn't know better, they are multiplier not target value.
eg: (FSB) 166Mhz base clock x 22 = 3.652Mhz‬ x2 = 7.304‬ GT/s of QPI bandwidth

So I have a pre-production DX58SO not capable to use the intended "Host Clock Frequency Override" variable.
Changing any valid value (240-133) make the board unbootable just after 136Mhz on any processor I've been able to test: i7920 - X5670 - X5675
To get to display any video output with an X5670 - X5675 on 5600 BIOS I also need to short the ▲12 J55EV  ▼21 J54EV ( to the non-default position ).

Is someone able to check on the 202 Board Reference Design something? or got the Allegro .BRD files for the DX58SO?
Thanks.

@Lost_N_BIOS got any idea?

@noInk CHECK this:D downside is RAM = low Frequency

@eivrah

Nice, I'm having such a bad experience with the inability to change the frequency on mine and knowing I won't be able to fix it duo some undocumented intel mistake.
You just look out for the temperature.

@plutomaniac,
@fernando
@lost_n_bios
is it possible to make a fresh bios of intel motherboards.
intel provides bios in 3 packages .bio .exe .zip
i extract thrice file one by one many MB files but not successed
bcoz after extact these .bio file we gat data in complex state means sometime
only header match some time footer match with original dump bios.
so i m not able to make fresh bios file.
if we have old bios dump we can clean me region, if problem is on me rgn side problem will be solve
but mostly time bios region corrupt.now we need only bios rgn from new bios but we are not able to extract
bios rgn in right way from .bio file
so finally i want to ask is any possibility to extract bios region from intel .bio or .exe files
i listen somewhere a tool (LBE) can make intel bios files but i did not tried it.
i upload a dq77cp bios file with exracted data too.
any solution to make bios region from it?

https://drive.google.com/file/d/15RHPKx8...iew?usp=sharing

@noink - sorry, not sure. Can you see/set PCI/PCIE Freq in BIOS, to lock that to 100? If not, confirm bclk 100-106 raises PCI/PCIE by using clockgen , if you can find crystal info on that board and it's in clockgen. This might be the issue.
Or, could be microcode too, did you try latest or is that not bootable with that BIOS? Some X58 have bclk holes too, try jumping around by 5-10 and see if that helps, ie instead of 200 try 195/205, instead of 166 try 151/171

@IntelModder - may be able to answer that for you @earthoo7

@fernando
@lost_n_bios

please share link which thread maching with this type problem.
& which thread should i can post this type problem.

ok lost_n_bios i got the link
thanks all of you