Thanks for trying to help me. I really appreciate your work.
Still doesnt wanna let me boot from the pcie slot. Things ive tried are:
flashed v10
Disabling the ASMedia SATA controller
Switching to nvmexpress_dxe_2 with no luck.
Installing windows 10 in uefi mode on a working system.
extracting and inserting nvme, nvmeSmm, NVMEINIT13 (says guid already exsists)
I am sure you already looked but just in case: Are you sure your SSD don’t appear in your list in your bios under an other name like PATA SS or something like that? It may be worth a look if you haven’t already. Sometimes it also takes a weird name like “Boot manager”.
PATA does come up but booing from it does nothing
This may be it. An other thing i read somewhere was one user who had a MSI motherboard and wasn’t able to boot with nvme because his bios was needing one slot of ram memory to manage it. He was able to use his disk only after freeing one RAM memory slot in his PC. If your PC has memory bar in all your memory slot then maybe, just maybe, it can be the problem. See this post for more details.
Also, if you have the “Boot manager” option in your list of disks, it may be worth a second look because, usually, all disks that are UEFI capable (usually GPT formated) are listed in the “boot manager” option.
I do have 4 x 4gb sticks. I removed the last one and it didnt seem to change anything
I forgot to say that in his case, the bios was trying to populate the DDR1 slot. So if you want to be sure that it’s not the problem, you have to free the slot 1 only, no other slot will do, assuming this is your problem. If you free the slot 1 and there is no change, you can be sure that it’s not the problem.
I removed DIMM 1 but still no luck
Ok, at least now you know for sure that it’s not the problem. So far i think your disk is detected by the bios because you see PATA in your list. Now we must figure out why it doesn’t boot from it.
Did you try to look for a “Boot manager” option in your disks list in the bios? Because usually any disk that is UEFI capable (Usually GPT formated) are listed in the “Boot manager” option. Usually any disk that are listed WITH and BEFORE you click the “boot manager” option is a link to boot from that disk in non-UEFI mode (like a MBR formated disk). My theory is that if you try to boot by clicking PATA without going into the “boot manager” first then it won’t work because the bios is trying to boot a MBR disk but yours is GPT formated to be UEFI compatible.
Sometimes, instead of being listed in the “Windows boot manager”, it is simply listed as a double entry: Like PATA and UEFI: PATA.
With the pcie card in I have Windows boot manager and PATA and without the pcie card I have just Windows boot manager. So that PATA option is definitely the pcie card.
Ive cleared cmos and there are no usb, hdd, or any other storage medium attached and I can still see Windows boot manager.
I have tried booting from both Windows boot manager and PATA and they both fail.
The other system that I tested the pcie card and ssd on is Asrock Z77 with modified bios and Ive install windows in UEFI mode.
So the disk is really there and detected by the module. But something is preventing it from booting. Would it be possible that the Asrock system is creating a boot file that is not compatible with your system? One way to find out would be to install Windows directly on your system by using a UEFI bootable USB key with Windows on it ready to install Windows in UEFI directly on your disk on your PC. I don’t see any reason why your system would not boot from it after that. Also do not forget to delete all shown partitions of the target Disk drive when installing Windows because GPT disk requires very specific requirements on the disk and the best way to be sure that Windows allocate each part the right way is to delete everything and give it only one big partition and let Windows manage it during the install.
BIOS with AGESA v1.5.0.7 updated to v6 Hybrid
Changelog:
[Changed] - NvmExpressDxE_2 -----> NvmExpressDxE_3
[Changed] - OROM LAN Intel Boot Agent GE v1.5.84 -----> v1.5.85
[Changed] - AmdProcessorInitPeim AGESA 1.5.0.0 -----> AmdProcessorInitPeim AGESA 1.5.0.2
[Fixed] - Increased integrity of the BIOS file. GUID DE3E049C-A218-4891-8658-5FC0FA84C788 file needed to be rebuild.
It is also fairly probable that the Cool and Quiet issue is fixed by this update.
The Hybrid part is this: Up to v5, this bios was, in a way, already an hybrid by the fact that the AGESA driver (AmdAgesaDxeDriver) was updated to v1.5.0.7 but the AmdProcessorInitPeim module was still in the original AGESA v1.5.0.0. With this update, the AmdProcessorInitPeim AGESA is updated to v1.5.0.2.
Phoenix48, I used F8 to get to the boot menu to select UEFI usb and I got it to install this time. I originally was looking for this option but it was never there and I made the boot usb the same way both times with diskpart. All in all, it works and Im really really greatfull to you. Thanks alot man 
Excellent! Enjoy your new system over pci-e!
hey.
v6.1a beta - cnq seems to not be working. OC seems to work.
v10 - cnq and oc working.
@pershoot
Thanks for the feedback.
Unfortunately, i won’t be near that PC untill the week-end but i will test that as soon as i can.
If i get the same results, then i can see only 2 possibilities why this could happen:
- Maybe the AGESA 1.5.0.2 is somehow not compatible with the rest of the bios for the Cool and Quiet function.
or…
- Maybe Cool and Quiet is affected somehow by the AmiAgesaDxe module that wasn’t changed in v6.1a. If it is that, then v6.1b should not have this problem.
CrosshairV-Formula-ASUS-1703.ROM with AMD AGESA OrochiPI v1.5.0.2 mod v6.1a (2018-03-04)
This version causes issues with the way AMD is reporting in Radeon settings.
https://snag.gy/MuoNVA.jpg
This bug shows with the AGESA updated versions. Also causes issues with loading some games, Fortnite hangs system on loads. Some system stability odds and ends as well, not sure if it’s to do with USB, SATA or Audio latency causing the issue.
I’ve got everything updated as far as it goes, USB firmware from your first post, AMD sata controller to 1.3.1.277 from this forum signed.
Tried looking into it with latencymon, looks like the network adapter on the board causes the highest execution, but the ISR count is the highest with HDAudBus and storport. So I’m leaning towards an issue with either the sata controller or Realtek causing issues with performance.
Windows 10 build 16299
Cheers
Thanks for the report.
I will look into that this week-end to see if i can observe the same memory glitch in my AMD video card.
As far as latency goes, you may try to switch to the 1.3.1.276 driver to see if it help. I was the one who provided the v1.3.1.277 driver for testing in this forum (extracted from an HP driver pack) and, even if i use it myself without problem, there was one report from someone saying that he was getting a lot of latency in latencymon with this driver. This seems to have stopped as soon as he put back the Micosoft version. I tried to test it myself with Latencymon but i had no problem with the v277 driver so i kept it running on my system ever since. I admit i was using the Bios with agesa v1.5.0.0 at that time and i didn’t test that again since but i doubt that problem had any relation to the bios because chances are this guy had a completely different motherboard and bios. You can read his report here.
The v1.3.1.277 is a driver made for HP for their servers so, in theory, it is optimised for that. And this also means, again in theory, that it may sometimes not be as much compatible and get along well with as many other devices in a system as the v1.3.1.276. So you may want to try that one to see if at least your latency problem go away.
As for the network adapter, it is normal because it is almost always the device with the most latency on any PC. This is the same results on all the PC i own. Of course, i am only talking about enough latency to get listed first in a PC. But if it is in first position by a crazy amount compared to the other devices, then there is a problem.
hey thank you so much for your V10 version i had received my samsung evo 960 and windows couldt install on it, i looked up to redownload the V9 version to reflash the bios to see if it works, and i saw that you update it to V10 with newer nvme support that is fantastic cause i have succed on installing windows on m.2, the only down side is the crosshair only has PCIE 2.0 and even on AHCI my speed is cut to half the predicted m.2 speed. but it still 3 times faster than a ssd so thank you very much that makes me very much happy thanks alot
CrosshairV-Formula-ASUS-1703.ROM with AMD AGESA OrochiPI v1.5.0.2 mod v6.1a (2018-03-04)
Was giving me crashes playing Overwatch for about 10 minutes consistently, could not get through more than one match without a complete system freeze. I want to say it’s because of a memory leak because the system locks up without a blue screen and must be hard reset.
Switched to
CrosshairV-Formula-ASUS-1703.ROM with AMD AGESA OrochiPI v1.5.0.0 mod v10 (2018-03-04)
Played 3 hours of Overwatch with zero crashes, also streamed so it looks like it’s super stable. I’ve had 24 hour up time as well without any issues so it looks like there’s something in the AGESA V1.5.0.2 mod that is causing issues.
LatencyMon is looking cleaner as well, more consistent execution time.
AMD sata controller 1.3.1.277 has been the most stable for me , 1.3.1.276 works well but doesn’t seem to have the same performance benefits. I’ll have to do some testing later to be certain but I’ve had better luck with 1.3.1.277.
Thanks again, I won’t be upgrading my system till probably late this year so all these bios improvements and drivers go a long way. Glad to do any testing you need me to just let me know.
After many tests with AGESA 1.5.0.2 v6.1a and v6.1b, and also AGESA 1.5.0.7 v6 Hybrid, here are my results:
- All modified versions of the AGESA have these 2 bugs (AGESA 1.5.0.0 v10 is the only one working correctly):
1- Cool & Quiet not working.
2- System memory amount is wrongly reported in AMD Radeon settings.
I was expecting this problem to NOT be present in the v6.1b bios because it is the only one that had its AmiAgesaDxe module changed to match the AGESA v1.5.0.2. But it makes no difference.
I think that something is not compatible in the bios between either:
1- Any new AmdAgesaDxeDriver AGESA version with another unkown (yet) module that need to be updated to match the new version.
or…
2- Any new AmdProcessorInitPeim AGESA version with another unkown (yet) module that need to be updated to match the new version.
For now, i don’t know which module is responsible for the Cool’n’Quiet management. If i can somehow find which one is, then i may try to update that module to match the new AGESA.
Those are all only theories for now…
I will have to do some deep thinking and probably do many tests to try to pinpoint the problem.