The first revision of the built in bioses have trouble updating from UEFI
im talking about gaming k3 motherboard.If you have it tell me how you config it becouse i not see any of the options i see on videos.
The K series boards are very poor and very basic
Hi Ket.
Your bios mod F9c for gaming 5 work very fine, thanks for your work. I can oc the corsair lpx 3200 hynix and the cpu 1700 up to 4050 with 1.42 vol stableā¦ not for 24/7 of course.
Do you taste the new Agesa 1071 for asus board? Some thing useful for us? Do you know some news about it?
Really thanks for allā¦
Not so, Iām currently testing the Biostar X370GT7 and its running my Samsung B-Die at 3200MHz perfectly fine, and 3333MHz as well as 3466MHz for that matter. It also doesnāt have broken 2400 and 2667MHz dividers like the Aorus boards do either.
I do know AMD have completely reworked the AGESA with 1071 (which I assume is actually meant to be put as 1.0.7.1). Asus seem to of jumped the gun a bit with the new AGESA but I have heard Asus boards have suffered quite a bit with stability issues so they are likely more eager to move away from early AGESA versions as quickly as possible.
My trident z 3200 ram is working at 3333 as well on my gaming 5
Yeah the Aorus Gaming 5 I have can run my test kit up to 3333MHz as well, 3466 is very flaky on the Gaming 5 but will boot at least. The 2400 and 2667 dividers being broken though on a board with what is already an extremely stripped down firmware is just inexcusable Gigabyte have no excuses for not quickly fixing issues like that.
For the difference in performance between 3200 and 3466 its not worth the hassle, 3200 is all i will ever need tbh. Really is pointless number chasing to be fair.
That depends on the engine and various other factors. Faster memory for Zen does have noticeable and beneficial advantages its not just number chasing. 
ket you are doing amazing work with bios.I was wondering if you modded bios can run RAM at 3200? (B350 Gaming 3)
I canāt reach it with the official one.
It might be a memory compatibility issue. In theory its something that is fixable it just requires identifying which part of the firmware stores that information, its probably got its own module. I havenāt really looked too deeply at this so donāt really know more than that and wonāt be able to test anything for some time where Iām testing a different board now, the Biostar X370GT7, it has its quirks, but its a much better board than anything Gigabyte are offering.
Everythings working perfect with this BIOS flash in windows, but when in the uefi BIOS itself the CPU ramps up to 60c, is there any way to stop this?
As far as Iāve seen thats normal behaviour on any AM4 system. You might be able to stop it by adjusting a few P-State options.
Any way to reduce the firmware load times further? I find they are very long 10+ secs atm. Any specific settings I can change to help with this. without disabling usb or Is that my best option? On your AX370 Gaming 5 modded UEFI F9c bios.
Ok cool, im aware thereās an offset of some sort making the cpu appear to be higher than it really is in temps so that the fans ramp up, on my old asus 970 aura motherboard there was a setting in the bios to stop this from happening.
Going back to the p-states thing, there are 2 settings, one for ZEN OPTIONS and the other somewhere else listing p-states 0 to 7, now ive managed to work out how to use the zen options part but in the p-states section, nothing can be selected?? well i can select 1 to 7 but they dont stay there, goes back to 0. hmm.
anyway in a couple of days im going to donate to you, because without this bios you have done this board is a PITA.
Are you using the reverse engineered UEFI? Theres two sets of P-State options in that, I wouldnāt recommend trying to use them both together.
Indeed, I commented on the Gaming 5 having quite good hardware but as you rightly say without good firmware to back it up the hardware is worthless and vice versa. I donāt like the vcore stability on the Gaming 5 though Iāve tested numerous AM4 boards now and the Gaming 5 has by far the worst vcore stability. I will never understand how Gigabyte broke the 2667MHz and 2400MHz dividers though that is just piss poor quality assurance. Neither the Asrock or Biostar boards Iāve tested broke those dividers. Gigabyte should be utterly ashamed that Biostar, one of the smallest manufacturers when it comes to mainboards at least not only manage to have better quality assurance but thoroughly embarass Gigabyte when it comes to firmware options and features. With some more testing Iām fairly certain I can get my test kit of Ripjaws on the Biostar GT7 to 3600MHz something I definitely canāt do on the Gaming 5 because if Gear Down is disabled the board wonāt POST. The GT7 also returns the highest benchmark scores of all of the AM4 boards I have so theres a lot they are getting right.
That sounds like the DRAM boot training slowing things up, nothing you can do about that. Buying RAM specifically for Zen might help.
"We skipped an AGESA (1071) because we felt it was too buggy for even BETAā¦ Good news is new new AGESA/BIOS (1072a) ETA is Friday"
From gigabyte forumā¦
I did say Asus jumped the gun a bit with AGESA 1071, but from what Iāve seen its working fine on Asus boards. Gigabyte probably needed the extra time to fix their own bugs and shortcomings in their firmware. Weāll find out soon enough.
Yes i am mate, using the reversed engineered bios from here, i do not understand how to use the second set of p-state options though, under ZEN OPTIONS got that part figured out, but the other set of options going from 0 to 7 i do not understand how that works. the vcore on my gaming 5 is actually running very very good indeed, on the stock spire cooler running at 3.7ghz at 1.285v fully stable with intel burn test AVX 164 GFlops
I donāt use P-State overclocking but ā0ā I think corresponds to the highest CPU frequency while ā7ā corresponds to the lowest frequency. Again I wouldnāt recommend trying to use both sets of P-State options at the same time one or the other. If you arenāt already Iād suggest at least using hardware monitor to check vcore stability. Under 100% load using the seti@home project vcore had up to a 0.061v variation even with LLC on the āHighā setting which is just awful. The latest board Iām testing which is a Biostar GT7 only varies by 0.02v and thats without any LLC.