Print Page | Close Window

New AGESA 1072a incoming!

Printed From: ASRock.com
Category: Technical Support
Forum Name: AMD Motherboards
Forum Description: Question about ASRock AMD motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=6805
Printed Date: 30 Dec 2024 at 3:23am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com


Topic: New AGESA 1072a incoming!
Posted By: SoniC
Subject: New AGESA 1072a incoming!
Date Posted: 03 Dec 2017 at 6:11am
" rel="nofollow -
Many sources cited some Gigabyte representative on why GB has not yet rolled out the AGESA 1071 in the newest UEFIs.
Answer was that they'll skip 1071 and go directly with 1072a as this one is far more stable.
ETA to release it to the board partners was this Friday (1.12.2017). 

Assuming everything went well and AMD delivered - we could see AM4 UEFIs with AGESA 1072a popping up in the coming 1-2 weeks.


Sources>
https://www.reddit.com/r/Amd/comments/7ggoml/new_agesa_1072a_eta_is_friday/
https://overclock3d.net/news/cpu_mainboard/amd_is_set_to_release_agesa_1072a_bios_updates_for_am4_motherboards_this_friday/1
https://forum.overclock3d.net/showthread.php?t=82637


I really do hope that some of the improvements will also be ported to TR4/SR4 X399 platform...

UPDATE:
According to Asus, CPU ratio OCing is broken in 1072a.
So probably 1072b it is ;-)

UPDATE2:
According to many ppl even AGESA 1071 does improve the memory compatibility (probably by fixing the cold bug 
(
described here ->
http://forum.asrock.com/forum_posts.asp?TID=6349&PID=39612&title=agesa-1007-incoming#39612
)
examples:
--People report 32GB of RAM at 3200 etc, things they couldn't do before.
--I'm on 3008 now (ASUS UEFI incl. AGESA 1071). No cold boot issues, holds 16GB@3200 CL14 with SOC@ 1.15V(auto) and 1.35RAM voltage 3,85GHz at 1.35Vcore. Stable. The only bug I found is that offset voltage doesn't seem to work, had to put in the value manually.

UPDATE3:
First GB/ASUS UEFIs are out. There are some OC bugs but what is hearth warming is this entry:
Last minute update from their forum: Quick update before the weekend: last round of testing identified a bug in 1072a. We just received the patch from AMD. They will be working on it first thing next week. Depending on testing it should be available early next week (assuming the patch works and no new bugs are found).

So I assume AM4 platform will receive some love next week, and if eveything goes well- in two weeks beta, 3 - official?

Still no love for TR4/SR4......Angry


-------------
-=SoniC=-
TR x1950, ASRock X399 FPG (v. 3.33A), G.Skill 3200 CL14 64GB, Enermax LiqTech 280,
AMD Vega 64 LC, 10x HDDs (mostly Hitachi), 2x M.2 (970EVO,960Pro), Xonar DX, AX1200i PSU



Replies:
Posted By: Maddreg
Date Posted: 05 Dec 2017 at 4:06pm
" rel="nofollow -
Few models already got BIOS with AGESA 1.0.7.2

http://www.asrock.com/support/index.asp?cat=BIOS

-------------
ASRock AB350M Pro4//Ryzen R6 1600@3800=1.275v+Macho rev.B//GTX 1060//Ripjaws V@3200 16-18-18-38-56=1.4v (2x8Gb)//Corsair RM550x


Posted By: foppe
Date Posted: 05 Dec 2017 at 5:43pm
After restoring defaults, updating to 3.30 and then 4.20, I can no longer apply manual OCs to my CPU, either via the main screen or the p-state OC menu.


-------------
AB350 Pro4, Ryzen 1600, 3.8GHz@1.35625v, Vengeance LPX 4x8GB b3000c15 2666@1.35v


Posted By: SoniC
Date Posted: 06 Dec 2017 at 3:17am
Originally posted by foppe foppe wrote:

After restoring defaults, updating to 3.30 and then 4.20, I can no longer apply manual OCs to my CPU, either via the main screen or the p-state OC menu.

Read my updates -- ASRock probably used AGESA 1.0.7.2(a) without noticing anything (that Gigabyte already found out).

Clearly stated that GB found OC not functioning anymore and they already received a patch from AMD (to fix that). Hopefully ASRock will address that soon too.


-------------
-=SoniC=-
TR x1950, ASRock X399 FPG (v. 3.33A), G.Skill 3200 CL14 64GB, Enermax LiqTech 280,
AMD Vega 64 LC, 10x HDDs (mostly Hitachi), 2x M.2 (970EVO,960Pro), Xonar DX, AX1200i PSU


Posted By: chainsaw
Date Posted: 06 Dec 2017 at 4:11am
so ASRock used the Bugged version of AGESA for the update. GG

-------------
ASRock Fatal1ty X370 Gaming K4 P3.40
AMD RYZEN 7 1700 @ 4Ghz 1.39v
Noctua NH-D15 "blocks top 1x pci-e"
16GB GSkill Ripjaw V F4-3200C16D-16GVRB @ 2933 14-15-15-15 T1 1.365v
GTX 1080 x2 Sli


Posted By: SoniC
Date Posted: 06 Dec 2017 at 6:26am
Originally posted by chainsaw chainsaw wrote:

so ASRock used the Bugged version of AGESA for the update. GG

yup, looks like it.
They probably will release a beta or the users would need to wait for an updated UEFI.

If you don't have any memory compatibility problems with the AM4 platform then I would wait.

Still hoping that once AMD is done with AGESA 1072 (let's call it "b") then they will start on AGESA 1.0.0.5 for TR and include some fixes in there too (cold boot memory bug, which might prevent mine from running full speed 3200 CL14 (currently only 2933 CL14 stable)).


-------------
-=SoniC=-
TR x1950, ASRock X399 FPG (v. 3.33A), G.Skill 3200 CL14 64GB, Enermax LiqTech 280,
AMD Vega 64 LC, 10x HDDs (mostly Hitachi), 2x M.2 (970EVO,960Pro), Xonar DX, AX1200i PSU


Posted By: SiX-P4cK
Date Posted: 06 Dec 2017 at 9:54am
Same problem here, can't o/c with the bios 4.2 on AB350 Pro

-------------
AMD R7-1700 @3.85Ghz 1.33V + Noctua NH-D15S
ASRock AB350 Pro4
G.Skill Ripjaws V Kit 2x8GB @3066 16-18-18-39
AMD Radeon RX 480 MSI GAMING X 4Go
Samsung 960 EVO 250GB SSD M.2 2280


Posted By: Maddreg
Date Posted: 06 Dec 2017 at 7:39pm
" rel="nofollow - I've seen several comments at Russian overclockers forum about broken processor OC on new AGESA, and today ASROCK rolled out same bugged BIOS for other model. I wonder why they are doing that?

-------------
ASRock AB350M Pro4//Ryzen R6 1600@3800=1.275v+Macho rev.B//GTX 1060//Ripjaws V@3200 16-18-18-38-56=1.4v (2x8Gb)//Corsair RM550x


Posted By: kerberos_20
Date Posted: 07 Dec 2017 at 2:02am
i have no issue with 32gb gskill cl14 at 3200 xmp, the only thing which prevented to boot em at this clock was invalid ddr voltage from bios (-0.1V real voltage ie 1.25 instead of 1.35)
https://imgur.com/cdcDugc" rel="nofollow - https://imgur.com/cdcDugc


Posted By: Spectre73
Date Posted: 07 Dec 2017 at 3:25am
Originally posted by kerberos_20 kerberos_20 wrote:

i have no issue with 32gb gskill cl14 at 3200 xmp, the only thing which prevented to boot em at this clock was invalid ddr voltage from bios (-0.1V real voltage ie 1.25 instead of 1.35)
https://imgur.com/cdcDugc" rel="nofollow - https://imgur.com/cdcDugc

Yeah. So what does this have to do with AGESA 1072a? You are using 1006b......


Posted By: kerberos_20
Date Posted: 07 Dec 2017 at 3:26am

and there goes 3333mhz..still on 1.0.0.6b tho :/
cl 16 sets on its own, cant change it to 15 :/


Posted By: datonyb
Date Posted: 07 Dec 2017 at 4:53am
you can change it but only if you disable geardown mode or bank group swap

cant remember which one disables 'odd' numbers

eg 13/15/17

also your latency is very high

you may well find much better overall performance on 3200 cl14 if you can get it stable


-------------
[url=https://valid.x86.fr/jpg250][/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold


Posted By: enhancer
Date Posted: 08 Dec 2017 at 12:39pm
I have found that the new 1.0.0.7 BIOS for my B350 Gaming K4(4.10) is preventing me from overclocking, and doing strange overvolting if I try. Normally, I get a significant Vdroop on my B350, requiring me to run rather high voltages to maintain stability. Now, it's doing the opposite in everything except Ryzen Master, i.e I set a voltage and that becomes the MINIMUM voltage it runs, pretty scary. I've tried flashing to previous BIOS versions and the problem remains.


-------------
Ryzen 5 1600
ASRock AB350 Gaming K4
G. Skill Ripjaws V 3600(3466Mhz 16-16-16-16-36)
EVGA GTX 1080ti SC Black
Samsung 960 EVO 256GB
Corsair Carbide Spec-Alpha
Windows 10


Posted By: kerberos_20
Date Posted: 09 Dec 2017 at 3:26am
Originally posted by datonyb datonyb wrote:

you can change it but only if you disable geardown mode or bank group swap

cant remember which one disables 'odd' numbers

eg 13/15/17

also your latency is very high

you may well find much better overall performance on 3200 cl14 if you can get it stable

i have fiddled a bit with subtimings and got stable 3333mhz cl14 :)



Posted By: datonyb
Date Posted: 09 Dec 2017 at 4:14am
well done


-------------
[url=https://valid.x86.fr/jpg250][/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold



Print Page | Close Window

Forum Software by Web Wiz Forums® version 12.04 - http://www.webwizforums.com
Copyright ©2001-2021 Web Wiz Ltd. - https://www.webwiz.net