[Guide] CoffeeLake CPUs on Skylake and KabyLake mainboards

Good afternoon guys, I need some help. I have a Maximus IX Formula, Z270, I made it the BIOS MOD using Cofee Lake 0.99 and Iā€™m using a QQLS. I recorded the BIOS via Asus Flash Back and with my G4560 it works perfectly but with QQLS the PC turns on error 00 and turns off. someone could help.

IMG_20220402_144309.jpg

Hi bf3itu,

quite funny. I m having the exact same symptom, although i think of me being a little further.
I modified my Bios with Coffee Time and tried the QQLS and a 9980HK. Both didnt work (00 and shutdown). I deactivated Intel ME using me cleaner and just got them to boot - unfortunately just for one time after a CMOS reset. When not being exactly after the CMOS reset there s the error 00 and the shutting down.
Ouh and for me: Its not an ASUS, but a MSI Z170 XPOWER (MS-7968).
Any hints for one of us?

Thanks in advance.

Hello everyone, this is my first post on the new forum and it looks wonderful so far. I come here to ask if anyone have experience on the mod working on the ASUS Z270-K? I have successfully made a successful mod on the Maximus VIII Ranger and presently running an i7 9700K at 5.1 GHz (previously tested working with a 9600KF and 9600K as well) and over time, I have made enhancements to the BIOS as well (latest F0 microcode, LAN ROM version to fix WOL, update ME to 11.7.0.1261 and adjusting default settings via AMIBCP to make Win 11 compatible by default).
I have already prepared the BIOS for the Z270-K. I donā€™t have the board yet but want to try to make sure if everything looks okay. Hereā€™s the link. Any help is welcome.

I have an Asus z270f gaming and want to know if this guide can get the board to use a 8400 (6 cores no HT)

Bios version: 1301
ME version: 11.7.0.1040

I know the board does not have flashback and only has crashfree3 (not sure if itā€™s useful or not)

Does anyone know if pin isolation is needed for this board? I plan to do find the pin 102 and ground it for the superIO.

Is there a way to flash the modded bios safely without bios programmer? If I screw up, only way to get it back is via bios programmer? I heard most of the programmer out there canā€™t do the needed 3.3v, 5v will kill the bios chip??? Any cheaper programmer with clip that works with 3.3v?

This forum has changed a lot since my last visit.^^
Could someone point out to me whether the same is true for CoffeeTime?
I would be especially interested in a download link for the most recent version which I fail to find in this new layout. :slight_smile:
Also, have there been further improvements regarding suitable UEFIs? I want to finally switch over my private VIII Ranger to 9900K use and last time I checked, there were still issues with IME and 16-thread compatibility in the newest ASU UEFIs, so use of older versions was recommendable. But there were already improvements over the initial ā€œ2202 onlyā€ limitations, which leaves quite a lot of possible candidates for best starting point.

Heyho Torsten,

latest Coffee Time can be found here (Post 1416 if link does not work).

Compatibility with Z170 is now nearly flawless. Running mine 9900k on Asus Z170 Deluxe with latest Bios 3801, all Cores working, HT working. OC to 5,2 allcore. Make sure to include the latest microcode with UBU to get rid of the AVX downclock bug. Without the latest microcode the clock does drop -0,1 Ghz on AVX loads.

Only problem I have is with the integrated TPM module, which does not activate, no matter what I tried. There is a relative long thread about that topic (couldnā€™t find it right now). On some boards it could be activated with lots of hassle, on some others it doesnā€™t. Bought an external module for 15ā‚¬ for Win11.

2 Likes

Thx a lot. TPM does not bother me and AVX will be power limited anyway (550 W PSU, fanless watercooling,GPU might draw 350 W => I will probably not go beyond PL1 150 W/PL2 200 W). But I will keep the issue in mind. More pressing though: To reach >5 GHz as well and to finally coax more than 6700K-compatible DDR3200 out of my B-dies. :slight_smile:

@svarmod or anyone else:
Are there any known limits to replacing IME with Coffeetime 99? Being on 11.8, I have to downgrade and I get shown the correct ā€œprocessingā€ and then ā€œReplace ME: Doneā€ when I try to do so. But than CoffeeTime reports the same 11.8.5 as before for the supposedly modified image. Some older modded BIOS images and .caps sourced directly from Asus work. (Both contain lower IME, though, 11.7. for the former; 11.6 for the latter.)

Also has anybody experience with drive recognition problems? After failing to modify my own file, I used the one provided by @itsakjt above ([GUIDE] Coffee Lake CPUs on Skylake and Kaby Lake motherboards - #1456 by itsakjt). But I got some UEFI-side initialization error when trying to boot from an SATA disk (being Win11, this might have some other problems with my system) and NVME disks where not even listed as boot devices, despite being accessabile as drive when running some live Linux. (I tried both my normal Seagate XF1440 U.2 and an M.2 image of the same system.)

Right now I have backflashed over itsakits file with some three year old 2202 based image made with Coffee Time 0.7, which, according to CoffeeTime .99 is missing half the patches. But at least I can boot into windows now. My SATAe attached front panel is not working, however, monitoring readouts are all over the place and donā€™t get me started on voltages. So I would really like to find sth better. :slight_smile:

(I CH143-flashed one of the UEFIs with uncessful IME downgrade but all other patches when still running the old CPU and back than everything worked fine. Except the 9900K obviously, so I guess a successful mod of my own BIOS image should solve most of the problems as neither IME nor CPU change should influence M.2./U.2 and SATAe of the PCH.)

@PCGH_Torsten - Which motherboard are you using? I have sometimes seen CoffeeTime failing to replace ME when the internal files of CoffeeTime are not renamed properly or are not present.

The matching board for your file upload ā€“ Maximus VIII Ranger. And I did not tinker with Coffeetimes files and obviously these are the same when I (successfully) mess with your file and when I fail to mod my own dumped BIOS.

1 Additon to the list of errors: System now failed three times in a row to shut off properly.

BTW @itsakjt: When running the same CPU on the same board (though with newer UEFI) as me,
did you observe any abnormalities with hardware monitoring and CPU settings?

  • I now use a -200 mV offset of the boards default 1.5 V Vcore (which do are not even close to any 1151 CPU) and I get 1.22 V under load and 1.32 V idle. Which is more akin to 1.3 V fixed preset on a board with really bad Vdrop than anything using an offset on actual turbo behaviour.
  • EIST and turbo modes seem messed up as well: My clocks vary between 2.9 and 3.6 GHz on all eight cores simultaneously.
  • Package power ainā€™t any better at 2.7 W acording to HWInfo 7.22, but the board does seem to know a more realistic value. At least my PSU reports 155 W on 12V1 which is about as close to my settings (150 W PL1) as above voltages are.
  • VR VCC temp is given at up to 93 Ā°C under load with a fast but rather small reduction on reduced CPU activity to mid-eightyish values. Normalle VR VCC should stay much closer to the temperature of its neighbouring MOSFETs, which are at a watercooled 45 Ā°C.

I am not even quite sure at the moment whether my readouts are trash or my board.

(Also my SATAe front panel now works for 30 to 120 seconds after powering on with drives vanashing in midoperation later. WTF? I guess will have to find that one other guy in the world that is acutally using SATAe.^^)

Hi @PCGH_Torsten,
I have upgraded the CPU now to an i7 9700K clocked at 5.1 GHz 24x7.
For your questions:

  1. I had to use an offset as well for that and it is actually quite common on ASUS boards. Do not worry about it. This was the case with both CPUs (9600KF and 9700K)
  2. I use a manual overclock as mentioned with quite high voltages (1.48V on load) and have left the EIST and turbo settings at Auto. They work perfect.
  3. I use AIDA64 for monitoring and it seems to report perfect values. For the vCore, I have measured on the vCore chokes and get same reading with multimeter.
  4. VR VCC temp for my unit is reported at lower 80 degree C on idle. I have researched on this and it is normal and is basically the temperature of the voltage controller inside the CPU. Do not worry about it as long as your CPU is cooled properly.
  5. As for SATAe, I have never used that on my board. I use a Samsung PM981a 1 TB NVMe SSD along with a 480 GB Kingston SATA SSD and a 2 TB 7200 RPM HDD from Seagate and they work fine,

Also later, I have updated the BIOS further more and fixed an issue with Wake on LAN not working (option was greyed out in UEFI) by updating the Intel LAN OPROM. I have also updated the CPU microcode to the latest F0 revision. I have also made changes via AMIBCP tool to make CSM disabled and secure boot enabled by default to make it Windows 11 ready by default. I also needed the ASUS TPM-M R2.0 card to have TPM support since the ASUS Z170 ROG boards typically have Intel PTT disabled through the FPF on the chipset (FPF = Field Programmable Fuse - Basically a one shot fuse). I spent 6 days on that and figured out that it is not possible to enable PTT on my particular board without physically replacing the chipset.

M8R_4502.zip (8.2 MB)

Please use AsProgrammer for flashing the BIOS with CH341a.
AsProgrammer_V2.1.0.13_XiTongZhiJia.zip (2.0 MB)

Please try this new file and let me know the results.

Also please let me know if you are setting your board data correctly (UUID, SN, MAC address) with Coffee Time.

Also, did you isolate any of the CPU pads? If so, you can safely remove them because this board do not have those reserved pins connected anywhere (my CPU socket is still in mint condition even after almost a year of heavy usage). Isolating the pads might have something to do with the false readings.

With the symptoms your PC is having, it looks like the BIOS is not getting flashed properly. AsProgrammer should fix it (select the chip model as on your board - typically Winbond W25Q128FV)

I still use this system as my daily driver and it still works great. I wish you will be able to run it perfectly as well. All the best and keep us updated.

Hi.
I was using SkyGz CH341 Programmer 1.4 free so far. If it did any errors during flashing, it sure made the same during verification. CPU is not isolated, just sktocc modded. (I aint that good and flimsy soldering, so modding the board seemed to be the less safe option.) UUID/SN/MAC were successfully transplanted into the BIOS I flashed based on your older file, with exception of the 2nd MAC that for some reason was found in my own dump despite there being no 2nd LAN controller on M8R. (It was identical to the 1st one anyway.)
AIDA package power makes a lot more sense, at least it goes up into tripple digits under load instead of going down as reported in HWInfo. However 114 W seem a bit low for an 9900K at 2.9 GHz and 1.22 V (which in itself is definitley too low for an 9900K given 200 W headroom) and would imply a terrible VRM efficiency, if the 152 W supply power reported by my PSU are correct.
I will try your new BIOS and software next, hoping for improvements. (Though I do not intend to let Win11 near my PC and in fact need CSM :-))

BTW: PC just did it first proper shutdown since switching the CPU. Well - too late now pal to show some good will! :stuck_out_tongue:

Nice work you did on your BIOS. Current LAN options, did even set the correct CPU voltage. But darn: Switching CSM back on did nothing for boot device recognition, my mbr windows installation remains unbootable on both NVME drives (tried M.2 and PCIe x4). An Win 11 GPT drive, however, is recognized. On 2nd thought this might have been the same problem I had with your older image, when I didnā€™t think to test any new installation specifically. Could it be that parts of your CSM hard deactivation overrule settings made in the BIOS?

BTW I just noticed your specific SSD: The PM981 was that one exception which brought its own option rom, making it bootable as an AHCI or PCIe device without need for functional NVME support, wasnā€™t it? I always wanted to try Win98 on one of those. :slight_smile: But more importantly it might be that your system skips the whole NVME recognition that could be involved in my problems.

I am now trying to advance my crude 2022 based attempt. With all CoffeeTime mods applied it did at least recognize an USB stick in the SATAe front panel as a boot device, though the very same drive was absent in windows 30 s later. More on stability, performance and shut downs (or lack thereoff^^) tomorrow.

Update on 2202 improvements:

  • EIST fixed
  • Turbomode stayed unavailable.
  • Shutdown problems rarer, but still there. Hang up often occured so late that windows did not remark on it in next boot. Once I got a bluescreen indicating power management - perhaps ACPI tables werenā€™t the only thing off?
  • USB/SATAe will work after deactivating and reactivating USB device in windows. Canā€™t be driver issue, though, because windows just sees a generic PCIe link and and ASMedia controller - the very same setup used with the very same driver for M8Rs own USB 3.1 ports without causing trouble there.

Next attempt: Got CoffeeTime99 to replace IME in Asusā€™ 3802.cap and backflashed that one. (still no chance of modding my original 3802 dump, though)

  • Vcore now too low, immediate errors in Prime95 and instability afterwards caused UEF corruption. Had to manually reset - and now feel confirmed in giving bonus for external reset switches (which M8R lacks) in my reviews.
  • +100 mV ended in 1.46 V in Prime95 and overheating.
  • +50 mV gives 1.42 V under load and 1.12 idle. Seems stable, but really: 1.37 without offset were insufficient for 4.5 GHz?? I had planned to overclock and the 92 Ā°C core temperature with 30 Ā°C water intake observed at 1.42 V are already unsustainable.
  • Turbo is still broken. While the system goes to 4.5 GHz now (3.9 before) all cores are still synced, making high single core clocks at reasonable energy conversion impossible.
  • HWInfo package power now moves in the correct direction and deviates only about 10% from AIDA. But both appear too low, indicating 190/210 W with 8 core Prime95, while PSU reports 310 W input into VRM. Considering the temperature problems, the PSU readings seem more realistic.
  • Front panel seems fine (for) now
  • Shutdowns: Weā€™ll see.

Hello Again Guys,

all done perfectly on intel 9600kf and z170a gaming m7 but the problem here is that the CPU clock is actually stuck on 0.8 and the XMP + game boost is not working, tried to overclock it manually but itā€™s the same,

not so stable, if I restart it, it might go for the windows or it doesnā€™t show anything and the Debug shows D6 error.

idk what to do to solve the clock issue and XMP

plz help!

<3 <3 <3 <3

EDIT: it reads the basic clock of the CPU 3.7 on the top right corner but if u look at the clock itself it show 0.8

Hello all,

Was looking to run a QQLS with my Gigabyte Z170- Gaming 3 but after reviewing the VRM spec started looking for a higher end board and found a good deal on an ASUS TUF Z270 Mark 1. My AliExpress seller informed me that on the Gigabyte board I would just need to flash the modified BIOS they provided but looking around here it seems Iā€™ll need a programmer. Any suggestions on which to purchase in US? Where can I find info on using the programmer on the TUF Mark 1?

Edit: Can I use my Rasp Pi?

Hello all,

I recently got an i9-9880hk mutant cpu from Aliexpress and I managed to get it working on a Asus Z170 D3 board.

I used coffeetime to patch the bios (I dumped it from the board using a CH341A programmer). I managed to get it working only with a specific ME version and then disabling that.
The version is 11.8.77.3664 Corporate, later I found out that starting with 11.8, coffelake CPUs arenā€™t supported on the board (if enabled).
Since I have access to the programmer, I thought it would be easy to change back. No dice.
Iā€™ve tried many combinations:
Leaving the original version 11.6.13.1212 and enable/disable.
Upgrading to many other versions said to be compatible with coffee lake cpus.

Iā€™ve tried using MEInject, to put back the older version into the modded bios (so that the filesystem state shows initialized - thinking itā€™s the same system)

Iā€™m thinking to use a supported CPU on the board with the original bios to see if that boots next, but this is more of a hail mary and would be quite annoying (needing to remove the spacers from the socket)

Does anybody have any hints/pointers?

I looked over many threads but the amount of information is overwhelming.

LE: I just had a look at the bios the folks from the Aliexpress shop provided, and that too has the same ME version.

@reactive Hi, I read your posts about G20 machines and found the creativity in the modifications on the two different G20 BIOS/UEFI exciting - thanks for sharing.

Having read it, I now have the impression that it might be relatively challenging to modify the manufacturer BIOS/UEFI exactly as you have done (because Iā€™m not sure which version itā€™s based on in your case, 1001, 1202 or 1402 and what other customisations may have been applied besides the CoffeeTime patch).

So I would like to enquire - might it be possible for you to upload the file of this functional version of the BIOS/UEFI (for the G20CI with initialisation of 16 threads)?

Since I actually have the same machine/mainboard type and would also like to install one i9-9900k (SRG19, R0), I think it would suit perfectly. The only changes I would make would be to the entries for the MAC address, serial number, etc. (of course you could replace these values with placeholders before uploading, for privacy reasons).