AM5 AGESA 1.0.0.7b

[ir_cow]

And the boot times? Any improvement?

If anything they are longer

 

[aanti]

This AGESA has fixed MCR, so you can boot really fast on stable configurations. You have options for thorough memory training.
If you keep FCLK in sync to UCLK, the latency hit is not that bad, though you miss a bit of bandwidth. Personally, I am quite excited that a config like this is possible.

 

[ir_cow]

It's starting to look like the X3D chips have better IMCs. Hicookie already broke 9000 a few days ago.

 

[TheLostSwede]

And Gigabyte has started to release the UEFI/AGESA updates.

The first boot takes an insane time now and the debug LED is stuck at 15 while this is going on.

All things identical, except the UEFI/AGESA version and latency dropped by over 2 ns.

 

[Gmr_Chick]

And Gigabyte has started to release the UEFI/AGESA updates.

View attachment 306337

The first boot takes an insane time now and the debug LED is stuck at 15 while this is going on.

All things identical, except the UEFI/AGESA version and latency dropped by over 2 ns.

View attachment 306338 View attachment 306339

Thank you for posting this! I have this exact board and it's good to know they've begun to roll out the update. I'm on F12a. Would it be worthwhile to update?

 

[TheLostSwede]

Thank you for posting this! I have this exact board and it's good to know they've begun to roll out the update. I'm on F12a. Would it be worthwhile to update?

I can't tell you what's worth it for you, but in my opinion it's worth it, although I also have a Ryzen 7 7800X3D in that board as you can see and the update is meant to improve the performance for X3D chips as well.

The new AGESA seems to bring overall memory performance/stability, not just for speeds over 6000 MHz.

 

[Mussels]

Running DDR5-5600 it seems.

Unrelated:
Something unexpected happened. This XPG 7200 kit has a EXPO profile, but if I run AID64, it restarts the computer instantly. But it passes y-cruncher and memtest5 (25 cycles). Kinda weird. Any idea?

I've seen someone else mention that on the web


Is it the benchmark that crashes or just aida? This other user had a bug with aida that caused hard locks

 

[ir_cow]

Is it the benchmark that crashes or just aida? This other user had a bug with aida that caused hard locks

Just AIDA64 when I do Read, Write or Copy for memory. Instantly restarts, MB does the post code start up and everything.

I just raised the CPU voltage and it works now. Strange that I had no problems with y-cruncher and 25 loops of Memtest5 before. Must hit the CPU differently I guess.

 

[Mussels]

Just AIDA64 when I do Read, Write or Copy for memory. Instantly restarts, MB does the post code start up and everything.

I just raised the CPU voltage and it works now. Strange that I had no problems with y-cruncher and 25 loops of Memtest5 before. Must hit the CPU differently I guess.

probably the speed at which it changes clock speeds or the number of cores used - a power balance issue.

 

[Deleted member 232534]

it's just good stability test,

i was able to make my ryzen 7700x -26 pbo "stable" in prime95 for hours, but it would give random error once a 2 days while gaming.

I tried aida64 test and it turned out to be only -8 pbo stable, but now it's really solid.

it also needs way less time to produce an error - it's suitable to determining the 'per core' pbo undervolt values - took me ~ 3 hours

unfortuanatelly its only 30 days trial

AMD Curve Optimizer any guides / experience

I don't want to mess with that, I just set limits to motherboard and that's it. Also not interested in using old BIOSes with bugs. I'm just curious why is significantly newer BIOS with numerous improvements over the course of months affecting performance in negative way. newer ≠ better, is the...

www.techpowerup.com

 

[Zanuzz0]

Hello guys!

Reading this post make me more confortable, but I still have something to ask u.

I've made a new setup for casual gaming and i'm facing a memory frequency problem.

I've bought a 6000mhz kit for a x670e tuf gaming plus wifi (ryzen 7 7800X3D), but i can't boot on a EXPO profile.

Once I enter on BIOS and set up EXPO I or EXPO II, the system reboot and i can use normally. But, after turning off the PC and turn it on again, I face a boot error, asking me to enter on BIOS and, when I do this, my memories are on 4800mhz.

I'm on the 1618 BIOS Version, and ASUS have provided a new version (1634), that updates AGESA to 1.0.0.7b version, but it's in BETA version.

Should I update for this version or wait to a stable version? Is someone facing the same problem that I'm?

 

[ir_cow]

@Zanuzz0 your not the first person to have problems with ASUS. Can you check CPU-Z and tell me under the memory tab if it's Samsung memory? Also the brand you bought.

As for updating or not. I would. Worse that happens is it still doesn't work for the memory.

 

[Gmr_Chick]

Hello guys!

Reading this post make me more confortable, but I still have something to ask u.

I've made a new setup for casual gaming and i'm facing a memory frequency problem.

I've bought a 6000mhz kit for a x670e tuf gaming plus wifi (ryzen 7 7800X3D), but i can't boot on a EXPO profile.

Once I enter on BIOS and set up EXPO I or EXPO II, the system reboot and i can use normally. But, after turning off the PC and turn it on again, I face a boot error, asking me to enter on BIOS and, when I do this, my memories are on 4800mhz.

I'm on the 1618 BIOS Version, and ASUS have provided a new version (1634), that updates AGESA to 1.0.0.7b version, but it's in BETA version.

Should I update for this version or wait to a stable version? Is someone facing the same problem that I'm?

Had this exact same problem with an ASUS X670E Hero, which, IMHO, was completely unacceptable. Especially when you consider that my kit was listed on the QVL for the board. Got so fed up with it that I returned that board and picked up an Aorus Master from Gigabyte.

The Master was able to both run and boot properly into Windows with EXPO enabled on the first try.

 

[Mussels]

Hello guys!

Reading this post make me more confortable, but I still have something to ask u.

I've made a new setup for casual gaming and i'm facing a memory frequency problem.

I've bought a 6000mhz kit for a x670e tuf gaming plus wifi (ryzen 7 7800X3D), but i can't boot on a EXPO profile.

Once I enter on BIOS and set up EXPO I or EXPO II, the system reboot and i can use normally. But, after turning off the PC and turn it on again, I face a boot error, asking me to enter on BIOS and, when I do this, my memories are on 4800mhz.

I'm on the 1618 BIOS Version, and ASUS have provided a new version (1634), that updates AGESA to 1.0.0.7b version, but it's in BETA version.

Should I update for this version or wait to a stable version? Is someone facing the same problem that I'm?

It'll 100% be related to your stock SoC voltage
Manually set 1.20v vSoC (AFTER enabling XMP, before rebooting) and see how it behaves, then try 1.25v and 1.30v if needed and see how they go for long term stability - you want it as low as possible, but obviously it needs to be stable

Stability issues tend to show up on cold boots, and at the transition from idle to load, especially lower load (like watching youtube videos, AIDA64 memory test seems good as a fast trigger, too)

The key is that different boards are less accurate than others, so 1.30v on your board could be the same as 1.25v or 1.35v on someone elses - without hardware probes, you can never know. This voltage varies between BIOS updates, but the voltage required varies wildly depending on how many RAM sticks you have, and your individual CPU


Remember that there is NO AM5 CPU that officially supports DDR5 6000 natively - it's overclocking on all of them. Boards that work out of the box just send higher voltages, out of the box.

 

[Deleted member 232534]

For me it didn't boot on stock XMP memory settings, while the previous BIOS did (1.26.AS01[Beta]).

it turned out i have to wait ~4,5 minutes for the first memory training. After training it boots instant.There are also new settings responsible for memory training and i've set them too.

i'm running my 5600c36 kit at 6000c34 1.35v now

 

[Zanuzz0]

Hi guys!

Can you check CPU-Z and tell me under the memory tab if it's Samsung memory? Also the brand you bought.

For sure! There it is:

The memories that i've bought are: ADATA XPG LANCER DDR5 32GB (2X16GB) 6000MHZ - AX5U6000C4016G-Dclabk

Especially when you consider that my kit was listed on the QVL for the board.

Mine as well, check:

Stability issues tend to show up on cold boots

That is the case that happens to me. If, in every single cold boot I made, I set up EXPO profiles on BIOS and reboot, I don't have problems. But, when I turn off the system and "cold boot" it again, the error shows up.

Manually set 1.20v vSoC (AFTER enabling XMP, before rebooting) and see how it behaves, then try 1.25v and 1.30v if needed and see how they go for long term stability - you want it as low as possible, but obviously it needs to be stable

About this topic, I'm a little afraid to do this and make some shit in my hardware, even cause I'm a noob at this things ahahha
If there are a walkthrough that I can follow up, it'll make me fell better ahahah

Reading this post makesa me confortable that I don't have a hardware problem in my setup, but I still want to solve this frequency problem, if possible.

About updating the BIOS to the BETA Version, I guess i'll do this tonight, and then try to set up EXPO setting and do my tests.. then I can return and tell u about the results.

After all this, I really need to thank u guys for this help and hope u can continue the job u all do helping others with their problems.

 

[Deleted member 232534]

Hi guys!
...
After all this, I really need to thank u guys for this help and hope u can continue the job u all do helping others with their problems.

please attach screens with voltages from hwinfo and zentimings like I did in my last post.

 

[ir_cow]

it turned out i have to wait ~4,5 minutes for the first memory training. After training it boots instant.There are also new settings responsible for memory training and i've set them too.

i'm running my 5600c36 kit at 6000c34 1.35v now

Sure is stable? Run like memtest5 or anything? The timings look fine, but your FLCK is 2066, but you never know if it good until tested.

 

[Deleted member 232534]

Sure is stable? Run like memtest5 or anything? The timings look fine, but your FLCK is 2066, but you never know if it good until tested.

yes i raised FLCK (infinity fabric) frequency too. This is stable tested in aida64 plus i'm running this for 3 days now in and its 100% stable.

 

[chrcoluk]

Just waiting for everyone else to release patch B So far publicly its only ASRock. I guess in the meantime I will use the Taichi for DDR5-7200 reviews. Dang. AMD just created more work for me lol.

Update: Did some A/B benchmarks with a RTX 4090. There is a good dip once you switch to Gear 2. Can't recommend going above DDR5-6000 (or your IMC 1:1 limit).

(2x16 GB) G.SKILL Trident Z5 DDR5-7200 F5-7200J3445G16GX2-TZ5RK (34-45-45-115-127)
(2x 16 GB) G.SKILL Trident Z5 NEO DDR5-6000 F5-6000J3038F16GX2-TZ5N (30-38-38-96)

LOW:
View attachment 305743
View attachment 305744

View attachment 305749
View attachment 305750

High:

View attachment 305741
View attachment 305742

View attachment 305747
View attachment 305748

View attachment 305745
View attachment 305746

You have mid range/low end boards in hand to test also? I remember you suggested to me clock wall was possibly down to board, and reviewers (and most overclockers) tend to have high end boards.

Gear mode best use case seems marketing and the easier ability to see if a clock wall external to the dimms.

 

[ir_cow]

I have a GB Gaming X A620M. Can't post 6000 with single rank without manually setting the SoC and VDDIO-MEM. Waiting for this newest BIOS to see if it got better.

Motherboard clock wall is real. Otherwise it wouldn't matter what you bought for a good memory OC. I think most of it comes down to trace length. Use to think layer count was a big factor, but 8-layer Gigabyte did better than 12-layer ASRock for Z690.

 

[chrcoluk]

I have a GB Gaming X A620M. Can't post 6000 with single rank without manually setting the SoC and VDDIO-MEM. Waiting for this newest BIOS to see if it got better.

Motherboard clock wall is real. Otherwise it wouldn't matter what you bought for a good memory OC. I think most of it comes down to trace length. Use to think layer count was a big factor, but 8-layer Gigabyte did better than 12-layer ASRock for Z690.

I can tell you I tried 3000CL12 on this board that my Z370 handled, and I seen something I never ever seen before, bios froze LOL, was thinking I am going to have to cmos reset this thing, but luckily on second post I quickly managed to flip settings back like a cheetah.

THe QVL lists still show reasonable memory speeds, but these QVL lists dont report the gear mode.

 

[ir_cow]

I can tell you I tried 3000CL12 on this board that my Z370 handled, and I seen something I never ever seen before, bios froze LOL

Happens to me a lot when pushing the mem OC.

 

[Zanuzz0]

please attach screens with voltages from hwinfo and zentimings like I did in my last post.

Sure, there it is:

I'll update my BIOS tomorrow to the beta version, installing AGESA 7.0.0.6b, and then i return with results.

 

[Mussels]

Hi guys!


For sure! There it is:

View attachment 306995View attachment 306996View attachment 306997View attachment 306998View attachment 306999

The memories that i've bought are: ADATA XPG LANCER DDR5 32GB (2X16GB) 6000MHZ - AX5U6000C4016G-Dclabk



Mine as well, check:

View attachment 307000


That is the case that happens to me. If, in every single cold boot I made, I set up EXPO profiles on BIOS and reboot, I don't have problems. But, when I turn off the system and "cold boot" it again, the error shows up.


About this topic, I'm a little afraid to do this and make some shit in my hardware, even cause I'm a noob at this things ahahha
If there are a walkthrough that I can follow up, it'll make me fell better ahahah

Reading this post makesa me confortable that I don't have a hardware problem in my setup, but I still want to solve this frequency problem, if possible.

About updating the BIOS to the BETA Version, I guess i'll do this tonight, and then try to set up EXPO setting and do my tests.. then I can return and tell u about the results.

After all this, I really need to thank u guys for this help and hope u can continue the job u all do helping others with their problems.

It could have been QVL'd on the older BIOSes that ran far higher SoC voltages, before AMD forced them lower
Some of these boards have a hardware offset, and they had to compensate by lowering not just their maximums, but their stock, XMP and EXPO values - and you betcha some of em were done in lazy ways without much testing (Found by the techtubers with hardware probes, so 1.30v BIOS was 1.350v reality - so they'd have to have XMP/EXPO max at 1.250v on those boards not AMD's 1.30v)
Throw in the laziness i've seen where they test on one high end board and copy-paste to an entire lineup (WHY IS MY BIOS FORCED TO RUN 16:10 resolutions, GIGABYTE?!?) and it fits that they've been doing things fast and lazy, wait for users to find the bugs and have some poor coder make quick changes assumung all the feedback is correct (and when it's not, you get QVL RAM no longer working)
Whats yours running at in Zentimings (and do let us know if its auto or something you manually set)

and beaten to it

1.02 Vsoc is quite low, tbh. That's peachy for 4800, but if it's not raising up then it wont do 6000 with that

Try 1.20v SoC (aka vsoc) and see if that works, if it does you should try for the FCLK at 2000 as well for better performance