erockerPretty sure I got this one last month from MSI?

Well ASUS had new fw with this 1.2.0.7 microcode also for last month.

TheOneI personally don't like Windows 11 because of the way the context menus are setup, always having to hit show more options.

That's the only issue i have with 11, everything else is genuinely easier to use.

Bruno_OWith Gigabyte you can save your settings to a file and import, which I did, but I also checked all settings to confirm they were the same.
After seeing the huge loss on 1.2.0.7 I've actually tested all Agesas / BIOS between .3 and .7, and with every update there was a slight drop in performance, accumulating to 2-3% from .3 to .7

I have an Arctic AIO 280mm and well ventilated case, and I always test after a fresh reboot + 5 min of idle time. Ran the tests multiple times and results were consistent.

It all makes sense as the 2 BIOS with the heaviest impact on performance had micro code security fixes, which I don't care since my Windows is clean.

At the end of the day, I have a 5800X (overclocked + old agesa) that's faster than the 5800X3D (locked + newer agesa) according to cinebench and cpuz results on reddit.

On my board 1.2.0.7 broke curve optimiser from working, its possible its glitched for other users too (honestly i wondered if a lock intended for the x3d was applied to the x)

Cr4zyIt could well be, but that's not something I can do much about other than pulling apart a full custom loop to do it and it would have to 100% work which I also cant test without a lot of effort.

The mini has had several updates none of them have ever mentioned or addressed the issue and as there's a decent sized community of people using it if it was a product issue im sure they would have been aware of it.

It still remains that the only fix that worked before AMD issued any BIOS updates was switching pcie4 to 3 and it's still the only way im solving the problem. I cant see it being anything but a hardware issue somewhere with AMD/x570/msi. For everything else it's almost flawless (because i dont use tpm and dont yet run w11), but its a pain.

Windows 10, AGESA 1206, Gigabyte Aorus x570 Elite. I have the USB issues on my HyperX Cloud Orbit S and Blue Yeti as well.

It's funny that if my HyperX headset is plugged into rear USB 3.0 ports, the audio stutters. Plugging it into USB 2.0 seems to bypass the issue.
If my Blue Yeti is plugged in rear USB 2.0 ports, my voice stutters. I have to plug it into a USB 3.0 port to "solve" the issue.

Bruno_OSame here with my 5800X + Aorus B550 Pro.
PBO +200, hitting 5.05GHz

On the latest Agesa I get 15650 points on Cinebench23 and 672 single core on cpuz.

Rolling back to Agesa 1.2.0.3 I get 16030 and 692.

My guess is that newer agesas have security fixes that impact performance.

Not sure because it was the same story with the first X370 bios 5 years ago (thats why I always upgraded mine to keep high memory freq)
You didnt bench a game or two to see the impact ? Might have been interesting

Seems to me it's getting worse. So, back to 1.1.0.0? Or even earlier ones (1.0.0.3 - 1.0.0.6)? Or I'm misguided in the AMD SAM usefulness on my system? Can't say I've noticed any improvement with SAM turned on TBH. Any thoughts?

VeseleiloSeems to me it's getting worse. So, back to 1.1.0.0? Or even earlier ones (1.0.0.3 - 1.0.0.6)? Or I'm misguided in the AMD SAM usefulness on my system? Can't say I've noticed any improvement with SAM turned on TBH. Any thoughts?

It depends on the game.

I see a nice boost in forza horizon 5 with SAM on at my default resolution of 3440x1440p on my system.

Halo infinite also gets a nice boost from SAM and I play both games on a regular so it stays on.

There are some older game where it does hurt performance like forza horizon 4 so one has to test it.

Bruno_OSame here with my 5800X + Aorus B550 Pro.
PBO +200, hitting 5.05GHz

On the latest Agesa I get 15650 points on Cinebench23 and 672 single core on cpuz.

Rolling back to Agesa 1.2.0.3 I get 16030 and 692.

My guess is that newer agesas have security fixes that impact performance.

the 5800x has had some settings for the 5800x3d applied to it
accidentally or deliberate, we dont know

JeagerNot sure because it was the same story with the first X370 bios 5 years ago (thats why I always upgraded mine to keep high memory freq)
You didnt bench a game or two to see the impact ? Might have been interesting

2-3% in synthetic benchmarks, that falls into the error margin of gaming benchmark... so didn't do it

Bruno_OWith Gigabyte you can save your settings to a file and import, which I did, but I also checked all settings to confirm they were the same.
After seeing the huge loss on 1.2.0.7 I've actually tested all Agesas / BIOS between .3 and .7, and with every update there was a slight drop in performance, accumulating to 2-3% from .3 to .7

I have an Arctic AIO 280mm and well ventilated case, and I always test after a fresh reboot + 5 min of idle time. Ran the tests multiple times and results were consistent.

It all makes sense as the 2 BIOS with the heaviest impact on performance had micro code security fixes, which I don't care since my Windows is clean.

At the end of the day, I have a 5800X (overclocked + old agesa) that's faster than the 5800X3D (locked + newer agesa) according to cinebench and cpuz results on reddit.

Bruno_OAt the end of the day, I have a 5800X (overclocked + old agesa) that's faster than the 5800X3D (locked + newer agesa) according to cinebench and cpuz results on reddit.

This is correct, anyone who had a 5800X knew that synthetic benchmarks were never going to be as good with the X3D, especially multi-core.
I tend not to build PC's for running Cinebench/CPU-Z benchmarks, but I do understand people like the 'bragging rights' thing with posting benchmark scores - an odd practice IMO, but each to their own!

Instead, I purchased the 5800X3D purely for gaming, and it's provided literally staggering results in many of the games I play that are very heavy single-threaded (racing simulations with heavy physics loads), for both the VR (Reverb G2 - really impressive results in VR with the X3D) and my triple monitors setup. I know that for some games the performance improvements were much less, but the X3D still beat the 12900KS in a 40-game round up, which is certainly not the case with the 5800X.
I'm running a 3080ti, 32GB DDR4 on a Gigabyte Aorus X570 Master mobo.

I gave the 5800X (PBO boosted to 4.97Ghz all core) to my kid to upgrade her 2700x/RTX2080 build since she plays at1080p monitor and it's been a nice boost for her - a 'win win'!

If you care about cinebench, get a 5950x or 5900x
the pressure for winning MT benchmarks is why intel slapped on those mostly useless E-cores

globespy - your system specs dont show the x3D, what GPU have you got with it? My experience going from 3700x to 5800x was mind blowing (~30FPS increase in CPU limited games) with just a GTX 1080 at the time, was it a similar change with your GPU? (whatever GPU it may be)

Found more details on the 1.2.0.7 performance bug: EDC bugs out on some boards if you PBO it above stock.

"The AGESA 1.2.0.7 PBO bug combined with a 65W TDP processor is hilarious: My 5600X is effectively stuck at 90A EDC"
People can always run stock EDC, or roll back one version if they werent affected by the fTPM stutter issue

Not everyone is affected either, i've seen reports of it happening on Asus and Asrock, but gigabyte being problem free.
I know the BIOS i'm on was simply changed from BETA to final without any changes to the file, so maybe asus goofed and used a slightly older version of it

What's EDC?

OctopussWhat's EDC?

a PBO setting.

Oh.
Glad I don't overclock then.

OctopussOh.
Glad I don't overclock then.

some motherboards default these to non-AMD values

Clearly with a system like yours you're not afraid of changing a BIOS setting, if needed.

Well I just set "limits" to motherboard and call it, that's basically all I do, aside from a few negative offsets.

My MSI X570 Tomahawk got this version in May:


MAG X570 TOMAHAWK WIFI | RETURN TO HONOR (msi.com)

Asus has not released any update for the Crosshair Hero VI and the Crosshair Hero VII, but they did for their strix MBs. They are taking more time that the other brands, i think they forgot that the users payed more for better support, not only lights and some bios options.

Even the Aorus X370 Gaming K7 got the update first. If i had known that Asus would give a support like this to the MB, i would have kept the Aorus.

I updated to this BIOS for my MSI X470 Gaming Plus and I have lost AMD Smart Access Memory. I know you are supposed to have an X570, Ryzen 5000 series and a Radeon 6000 series but it was showing it was enabled before.

Has anyone else lost support for S.A.M?



I hate having to revert to an older version of a BIOS as it loses my fan profile every time.

ZilaosI updated to this BIOS for my MSI X470 Gaming Plus and I have lost AMD Smart Access Memory. I know you are supposed to have an X570, Ryzen 5000 series and a Radeon 6000 series but it was showing it was enabled before.

Has anyone else lost support for S.A.M?



I hate having to revert to an older version of a BIOS as it loses my fan profile every time.

you have to enable REBAR in your BIOS, it doesnt default to on

Musselsyou have to enable REBAR in your BIOS, it doesnt default to on

It was enabled. It was still my mistake I was supposed to restart the computer after I enabled it like the prompt that came up told me to.



Thank you for the help and the quick reply!

ZilaosI know you are supposed to have an X570, Ryzen 5000 series and a Radeon 6000 series

Not true. My system is the example. The AMD official page has incomplete and false information.

OctopussWell I just set "limits" to motherboard and call it, that's basically all I do, aside from a few negative offsets.

some motherboards set those limits to infinity and beyond

VeseleiloNot true. My system is the example. The AMD official page has incomplete and false information.

I think they officially supported their latest generation but it must be up the motherboard manufacturers to implement. The performance gains are real when SAM is enabled and I'm happy it's working again. Wait do you mean you have the latest generation and it doesn't work?

ZilaosI think they officially supported their latest generation but it must be up the motherboard manufacturers to implement. The performance gains are real when SAM is enabled and I'm happy it's working again. Wait do you mean you have the latest generation and it doesn't work?

No. What I'm saying is that SAM support even older hardware than AMD specifies. You can see my specs on the profile. People report even 390X and Zen 1 enabling the feature. Seems that AMD lied, and there is no physical limitation against the SAM support, only software. I guess it's all up to motherboard makers and their BIOS implementation of AMD AGESA 1.1.0.0.