Nothing new in Windows 10 build 20190.
AMD cpu00800F12_ver0800126C_2020-06-19_62745CFB
Intel cpuA0661_plat80_ver000000C8_2020-01-29_PRD_5455215F
cpuA0661_plat80_ver000000C8_2020-01-29_PRD_5455215F.rar (90.3 KB)
!New_cpu00A20F00_ver0A200025_2020-01-21_E3C8C569.bin
!New_cpu00A20F10_ver0A201005_2020-06-26_689BC5A3.bin
Not sure if already submitted? (using MC_Extractor_v1.46.1_r156)
ASUS X570-E 2606 BIOS https://www.asus.com/au/Motherboards/ROG…/HelpDesk_BIOS/
!New_cpu00A20F00_ver0A200025_2020-01-21_E3C8C569.zip (5.67 KB)
!New_cpu00A20F10_ver0A201005_2020-06-26_689BC5A3.zip (5.67 KB)
Thank you @clackersx for the Zen 3 microcodes. They are new, not in the DB, because unfortunately MCE was not able to detect Zen 3 microcodes until a few days ago. That means that we have missed any Zen 3 microcodes that have appeared early these past few months. The ideal thing to do is to go back and re-download/check the BIOS images released in the past few months to get what we missed but that requires time. These 3 were added in the db today from that one ASUS mobo but there are bound to be more:
AMD cpu00A20F10_ver0A201005_2020-06-26_689BC5A3
AMD cpu00A20F10_ver0A201003_2020-05-06_EDC5C6C0
AMD cpu00A20F00_ver0A200025_2020-01-21_E3C8C569
@Marvin @westlake Do you guys keep the Microsoft dll from older builds? They may also include AMD Zen 3 microcodes that we missed.
I think a 3950x uses the F10 microcode. Is this correct?
Oh wait, I’m thinking of the cpu00870F10. 
I ran a few mcupdate_AuthenticAMD.dll’s that I have (starting from build 20150 from 13th of June until the most recent build 20190) with the new MCExtractor 1.46.1 and found nothing to report.
So MS hasn’t added Zen 3 yet at all, good to know. Thanks for looking Marvin! 
These microcodes
were added by many manufacturers back in early July. Then there was still the emergence of new AGESA.
If analyze the "List supported CPU", then you can see the addition of models 3600XT and 3900XT.
Also on the 20th of July, AMD announced graphic models PRO 4x50G and 4700G/GE. (OEM)
I do not know if they are of interest
!New_cpu006EC_plat10_ver00000051_2006-02-06_PRD_719E24F7.rar (4.12 KB)
!New_cpu30661_plat02_ver00000108_2011-08-05_PRD_41EF8BAD.rar (5.24 KB)
!New_cpu30661_plat04_ver00000108_2011-08-05_PRD_A2383900.rar (5.24 KB)
!New_cpu30661_plat08_ver00000108_2011-08-05_PRD_4A9BCDCE.rar (5.23 KB)
!New_cpu30678_plat02_ver00000806_2013-11-26_PRD_5FB9F4FC.rar (50.2 KB)
!New_cpu00A20F00_ver0A200025_2020-01-21_E3C8C569.rar (5.66 KB)
!New_cpu00A20F10_ver0A201003_2020-05-06_EDC5C6C0.rar (5.66 KB)
!New_cpu00A20F10_ver0A201005_2020-06-26_689BC5A3.rar (5.66 KB)
Yes, that makes sense. I have checked ASRock and Biostar for now and Marvin said that MS has nothing new. So ASUS, MSI & Gigabyte remain from the main players. If anything extra is found, it will be included in the next DB update.
Most of them are indeed (except the known Zen 3). I’ve added them to the next DB, thank you.
This one is actually two 0x800 microcodes in one file. Both of them are already in the DB. However, the 1st microcode is incomplete (2KB instead of 4KB) and someone fixed its Checksum to match the incomplete data for some weird reason so that’s why MCE said it is new. That is a mistake of course. At the next version of MCE, it will be able to detect and report incomplete/truncated Intel, AMD, VIA and Freescale microcodes.
Oh, wow, that’s impressive. Thank you [P3D] Crashtest! 
Nothing new in Windows 10 build 20197.
Here are some more
This makes no sense, how are there so many different versions of the same microcode revison?!
I do not know I leave it to someone who knows better therefore I posted them here
cpu00610F01_ver0600111F_2018-03-05 - 4 files
cpu00630F01_ver06003109_2018-02-27 - 5 files
how to determine really the latest version?
Edit:
Probably have to focus on the post-build Windows 10 Insider?
The last build has these microcodes and with what checksum?
I guess we need a discussion about this. I did a few test editing AMD microcodes and if one changes the data in the AMD microcode file one can just put random data there and get a new checksum and a "new" microcode. It is not as protected as Intel where the checksum contained in the header covers the full microcode. So technically one could generate unlimited numer of AMD microcode files that MCExtractor asks to report.
I guess for AMD microcodes we need a "source log" moving forward, meaning where was the microcode exracted/obtained information is kept as metadata in MCExtractor/this thread?
As a proof of concept here is 3KB of RANDOM data with the first 20 bytes replaced with the AMD microcode header.
DO NOT USE THIS FILE, IT IS RANDOM DATA!!!
!New_cpu00630F01_ver06003109_2018-02-27_02BB164E
Again, this is just random data with AMD microcode header.
!New_cpu00630F01_ver06003109_2018-02-27_02BB164E.zip (2.76 KB)