Any chance you can mod these features into the AB350N-Gaming?
Bios UI layouts are the same.
Oh dear sorry to hear that, but itās to be expected. Gigabyte used to be ok but iām talking in the days of the AMD Athlon XP, theyāre certainly not any good any more. I had the gaming5 pumping 1.85v into my other ryzen7 1700 needless to say it didnāt last very long. You ought to see how much hate there is for this board on gigabytes forums, its cringeworthy. i find it absolutely unbelievable that the last stocks of the gaming5 are now selling for more than a Taichi on ebay. so my advice to everyone owning this POS, put it on ebay right now before you lose your CPU and while prices are high.
Gigabyte need to completely rebuild the firmware from the ground up. The only small chance there is of this happening is for people to complain in the Gigabyte forums and send support tickets. A lot of whats wrong with the board is software related and easily fixable if Gigabyte get off their high horse.
Does anyone have time to test an experimental Gaming 5 firmware Iām working on? Primarily memory compatibility needs testing.
Hi ket. I registered only to give you an answer. Firstly sorry by my English, I am a Spanish guy.
I flashed the official BIOS f22 version from you f10 modded one, and I have some unexpected blue screen in windows 10 while playing games.
With you modded f10, Mem can be set to 3066 Mhz on my PLX Vengance 3200Mhz.
With the official f22 increased to 3200 but was unstable.
Now I give a try to your new modded BIOS f22d and itĀ“s amazing. AMD 1600X overclocked to 3.925 without problem and memos at 3200Mhz. No reboot yet.
So from my side, thanks you so much Ket.
That improvement must be due to some changes I made to the default memory configuration settings, good to see it working out in realtime as it does in theory 
That improvement must be due to some changes I made to the default memory configuration settings, good to see it working out in realtime as it does in theory
Doesnāt work for me though when comparing to the older F9 series. I updated from modded F9 to modded F22b and canāt boot at 2933 with XMP enabled anymore (still havenāt tested without yet), even when applying my old custom timings that were stable. LPX 2x16GB 3200 kit.
Did they screw up XMP in the latest version?
EDIT: looks like they did. Disabled XMP and everything is exactly as it was, except that now 3000MHz is an option if I loosen some timings (probably not worth it). Testing stability again just in case.
That improvement must be due to some changes I made to the default memory configuration settings, good to see it working out in realtime as it does in theory
Doesnāt work for me though when comparing to the older F9 series. I updated from modded F9 to modded F22b and canāt boot at 2933 with XMP enabled anymore (still havenāt tested without yet), even when applying my old custom timings that were stable. LPX 2x16GB 3200 kit.
Did they screw up XMP in the latest version?
EDIT: looks like they did. Disabled XMP and everything is exactly as it was, except that now 3000MHz is an option if I loosen some timings (probably not worth it). Testing stability again just in case.
Yeah Gigabyte are historically know for making things worse and not better. The trend continues with the latest firmware as far as memory goes Gigabyte have completely and utterly borked XMP. I havenāt fully tested but I can say that with XMP enabled even 2133MHz wonāt boot anymore if it wasnāt so sad it would be laughable. More annoyingly I have a good idea what Gigabyte have ballsed up but some modules appear to be protected so I canāt replace them.
Hi ket.
I have a few questions.
Could it be possible to enable Ryzen to continue power save mode during overclock or this is internal in the cpu?
Also the minimum voltage could be set by the user(it default to 400mv minimum in C-states at stock for my Ryzen 5).
Can the post time be made shorter?..like in the Asrock Taichi with minimum post time(no training, all compatibility and legacy off) of 4 seconds?
Thank you.
Facedesk Iāve just spent the last two days working on detailed firmware improvements for the Gaming 5 and K7 and now there is an official F22 firmware for the Gigabyte boards updating the AGESA to what I was initially waiting for but as it looked like it wasnāt going to come anytime soon started modding F22b firmwares. Needless to say there will now be a delay as I have to redo everything (which hopefully wonāt take as long) and of course it will be updates for every board as its a new AGESA. I was also working on a new reverse engineered K7 firmware for the Gaming 5 butā¦ poof goes all of that work gotta start over.
Sit tight folks, Iām on it.
The exposed/unlocked settings of the modded firmware images I make should cater for most of your needs. As I have said before Gigabyte REALLY need to trash the current firmware and start over its a complete mess in a lot of ways. I just make it less shitty.
Theres no excuse for this piss-ant companies lack of support, i feel sorry for you guys, asrock have released 3 biosās this year already for the taichi https://www.asrock.com/mb/AMD/X370%20Taichi/index.asp#BIOS just as i though things couldnt get any better my corsair rgb ram is now 100% stable at 3600MHz and beyond with new agesa 4.4 bios, thus making video rendering as fast as a bugatti veyron.
Gigabyte released AGESA 1.0.0.1a for gaming K7
https://www.gigabyte.com/us/Motherboard/ā¦v-10#support-dl
Can you update the custom bios for this motherboard?
The F22 firmware, quite astoundingly, is actually quite good in terms of squashed bugs and issues that have been ironed out. Lets not get carried away though theres still a fair few things that can be improved and bugs still exist in the vanilla firmware that have been there for a year now that Iām still having to use workarounds for in my modded firmwares. Gigabyte should not be being praised for functionally bringing their boards up to the level all other manufacturers managed a good SIX MONTHS ago. A large amount of these improvements also appear to be due to the new 1001a AGESA so work AMD have done, not Gigabyte. Theres also still many options that are hidden from the user that shouldnāt be other manufacturers have had them exposed from day 1 the only reason I can think of why Gigabyte are not exposing the options is simply where they hate true enthusiasts and really canāt be bothered with doing what they are paid to do and fix bugs. Thus it is upto people like me to expose options and fix as many bugs as possible.
Taking only the vanilla firmware in to account here I can (so far) confirm the following bugs and issues are fixed;
Mouse cursor is no longer set to some very low DPI setting in the UEFI it is now much more responsive and fluid.
Polling rate for voltages under the PC health status stuff has been significantly reduced so you can monitor voltages much more in realtime
Vcore stability is improved, still not great, but improved so there must have been some recalibration at least
Finally SOC LLC is fixed
At long last the Gaming 5 appears to be running my RipjawsV test kit at 3466MHz stable. Asus X370-F did this out of the box, but at least Gigabyte have caught up here.
I have numerous other issues and bugs to verify present or fixed but F22 is certainly a step in the right direction.
@persano yep working on it this is the AGESA I was waiting for it significantly improves memory compatibility and a slew of other fixes. This is the first AGESA I can confidently say makes the AM4 platform feel like something people should have got a year ago there were far too many bugs in the beginning and the platform just wasnāt ready. Iād guess AMD would have liked to held back on release at least a further 6 months to develop the AGESA but their dwindling money and resources at the time simply didnāt allow for that which is understandable.
F22M-D for the Gaming 5 is live, the rest will follow as quickly as I can make them.
All X370 firmwares based on F22 are now live, early versions released yesterday have also been updated fixing minor bugs and issues. B350 versions will follow in the next day or two.
ket,
I just tested you new F22M-D for the gaming K7 bios.
At first glance i face the following problems(they do not appear in the official f22 firmware):
-no boot after msdos flashing using the provided utility, code 50. I used the dual bios functionality and recovered using Q flash.
-p states do not work ok, changing to custom any p state besides the first one results in code 50 error
-any restart from windows (10) locks the board into code 0d, even reset button can not restart it, only power off can.
-gpt usb stick doesnāt boot (it seems it reads but very slow )
cpu: Ryzen 5 1600, mem:Corsair RGB 2x8Gb 3Ghz cl 15
Thank you for the work!
Try switching DIMM banks for your memory a lot of those no boot situations sound like a compatibility issue or something in the system that is unstable particularly with POST code D0 and 50, with the former being a CPU initialisation issue and 50 being a memory initialisation issue. Theres no reason why P-States wouldnāt work with the modded firmware I havenāt made any changes to P-States. Did you do a full CMOS reset after flashing? (remove battery and power cord, press power button 3-4 times). While there shouldnāt be anything wrong with flashing from pure DOS you might want to try using Q-Flash GB might have made some changes Iām unaware of that the DOS flash method isnāt entirely happy about. I wouldnāt rule out there being an issue with your USB stick either Iāve had a lot of sticks and SD cards suddenly start performing very poorly in the past - a sure sign the media is as good as fubard. So Iād say its worth using a different USB stick as well to rule that possibility out.
In the meantime Iāll check the firmware image but Iām 99% certain itāll check out ok but always worth checking.
EDIT: No problems I can see or find with the firmware image. Are you using any kind of TPM?
ket,
I am using no tpm. Usb stick had no problem booting with original bios after trying the modded one. I put your bios on backup and i am switching between them to test.
In fact I am using the firmware posted by gigabyte on the site before, the f22b. It seems they changed it and i will have to retest with that too, maybe the new problems are from there.
I will also do the full cmos reset and retry with the new version, i used the first one you posted yesterday morningā¦
Later edit:It seems the restart is a bug in the new firmware F22
http://forum.gigabyte.us/thread/739/offiā¦rs-club?page=18
Hi all has anyone had any problems with booting from pcie sata card after latest bios flash I canāt see the drives connected to sata card in the bios but i can see them in Windows so currently cannot boot my Linux partition I would like to know if maybe there is a bios option in the new bios that I have to turn on I would like to add that I reflashed my bios to previous version and I can see the drives
ket,
I repeated the flashing proces this time reseting the settings taking out the battery si shorting the pins.
Dos flashing gave the same code 50 error.Q-flash worked but the 0d code after restart command from windows remains. I also tested the f22 bios and for me is not giving the od error, i can not confirm the bug at this time.
xrelic,
i only have pciexpress nvme, and no, i had no problems with bios f22 in this regard.