AMD Readies AGESA ComboAM4 1.0.0.3ABB, Addresses Several Issues Affecting 3rd Gen Ryzen

[btarunr]

AMD today addressed multiple issues with its 3rd generation Ryzen processors through a highly-recommended update to its Chipset Driver software. To begin with, it stated that several users noticed anomalous behavior with 3rd generation Ryzen chips where the voltages and clock-speeds would be raised as the processor would misinterpret low-scale performance requests from certain software as a request to unlock higher performance states (combinations of higher clock-speeds and voltages to support them). This first came to light when users reported abnormally high voltages at idle when performance was measured by certain software that caused the Observer Effect. The new version 1.07.29 of AMD Chipset Drivers refine the AMD Ryzen Balanced Windows power scheme to be more aware of low-priority workloads and ensure the right state when the system is idling. AMD recommends Ryzen Master software. Version 2.0.0.1233 (or later), as they have fixes to the hardware monitoring module.

The AMD Chipset Drivers 1.07.29 also includes a "beta" fix for the bug that rendered "Destiny 2" unplayable on machines powered by 3rd generation Ryzen processors. The company had earlier tried to fix this bug through an update to its AGESA processor microcode, through ComboAM4 1.0.0.3ABA, although that particular version, not to be confused with the widely circulated 1.0.0.3AB, was found to be buggy and pulled. AMD said it's working on a newer version of AGESA, version ComboAM4 1.0.0.3ABB, which will include "a more comprehensive solution" to the bug affecting "Destiny 2."



AGESA ComboAM4 1.0.0.3ABB will also address a separate issue AMD chronicles as the "Event 17, WHEA-Logger" Windows Event Logger bug. Apparently, some users with M.2 PCI-Express SSDs installed on machines with 3rd generation Ryzen processors are experiencing a number of "Event 17" errors in Event Logger related to their boot drive. One user even complained of System File Checker reporting system-file inconsistencies, although AMD categorically denied accusations of "data-loss" on its platform.

AMD is currently testing and validating AGESA ComboAM4 1.0.0.3ABB and will push it to motherboard manufacturers to encapsulate into their BIOS updates in the coming weeks.

View at TechPowerUp Main Site

 

[hzy4]

Where can I get AMD Chipset Drivers 1.07.29, on the AMD Homepage the latest driver is 1.07.0725, and google search for it revealed only this article

 

[W1zzard]

Where can I get AMD Chipset Drivers 1.07.29, on the AMD Homepage the latest driver is 1.07.0725, and google search for it revealed only this article

https://download.amd.com/drivers/amd_chipset_drivers.exe

Not sure why they didn't put a version number into the filename..

 

[Mistral]

Good on patching these so quick, but some would say Destiny 2 not being playable is a feature...

 

[silentbogo]

Just give it another few days and ABBA is going to be back as well)))
BTW, I'm not sure whether the same issue is applicable to older Ryzen chips, but I've noticed that after upgrading to 1.0.0.3AB my 1600X became much hotter. Checked core voltages, and to my surprise it was boosting over 1.5V vCore during low-thread loads (that's with "normal" BIOS setting, not "auto"). On older version I've never seen such temps or voltages above 1.425V.

 

[danbert2000]

It's always a bit unnerving to see such serious bugs in a CPU launch. I guess the game bug isn't affecting too many people, and the power plan can always be tweaked to be more aware of edge cases. But these are issues with underlying causes that seem pretty serious. I read that the Destiny 2 problem was because it was using a specific instruction that has a bug. That's not good. And the fact that too much voltage is being pushed to processors near idle, how are these problems going to impact people that can't or don't update their software? This kind of confirms that, out of the box day one, Ryzen 3 was broken. Hope they fix it soon.

 

[shmuck]

https://download.amd.com/drivers/amd_chipset_drivers.exe

Not sure why they didn't put a version number into the filename..

They do use the version number in filenames, so I don't know where you got that link or what version it possibly is.

EDIT: The link is from https://community.amd.com/community...te-5-let-s-talk-clocks-voltages-and-destiny-2

 

[W1zzard]

They do use the version number in filenames, so I don't know where you got that link or what version it possibly is.

EDIT: The link is from https://community.amd.com/community...te-5-let-s-talk-clocks-voltages-and-destiny-2

Yeah, same link, no version number

that's the version info

 

[ironwolf]

OK, the amount of letters after the version number is getting comical.

 

[moob]

Installed both. The links are in the OP though neither has the version numbers in them for whatever reason.

Definitely seems to have addressed the fluctuations in voltage/temperature. And while I have Destiny 2 (they gave it away for free), there's no way I'm reinstalling that to test it. lol

 

[xorbe]

OK, the amount of letters after the version number is getting comical.

Could they have not just used 1.0.0.4 ...

 

[W1zzard]

Could they have not just used 1.0.0.4 ...

or 1.04, or 2.0 ..

 

[Assimilator]

https://download.amd.com/drivers/amd_chipset_drivers.exe

Not sure why they didn't put a version number into the filename..

I f**king HATE websites that fail to do this!

 

[bug]

AMD Readies AGESA ComboAM4 1.0.0.3ABB, Addresses Several Issues Affecting 3rd Gen Ryzen

followed by an article about a new chipset driver and

AMD said it's working on a newer version of AGESA, version ComboAM4 1.0.0.3ABB, which will include "a more comprehensive solution" to the bug affecting "Destiny 2."

Smooth.

 

[OC-Ghost]

Very nice!

But seriously just use numbers...
Summit/Pinnacle AGESA never went above 1.0.0.7, and I have a hard time imaging they effed up so hard they need to go over 1.0.9.9

 

[Imsochobo]

It's always a bit unnerving to see such serious bugs in a CPU launch. I guess the game bug isn't affecting too many people, and the power plan can always be tweaked to be more aware of edge cases. But these are issues with underlying causes that seem pretty serious. I read that the Destiny 2 problem was because it was using a specific instruction that has a bug. That's not good. And the fact that too much voltage is being pushed to processors near idle, how are these problems going to impact people that can't or don't update their software? This kind of confirms that, out of the box day one, Ryzen 3 was broken. Hope they fix it soon.

Let me address them in a reasonable unbiased way.
Destiny 2 is a random number generator instruction which hardly anyone use.
The instruction bug, very valid and it's been a long time since an architecture haven't introduced bugs like this.
Kl, cfl are refreshes no change so not valid.

Whea bug should really have been discovered and be fixed before launch.
Idle power, voltage, frequency is just kids screaming in fear while everyone says high number, low power consumption in 95% of all cases.
It's still a bug, but blown way out of proportions and I am one affected and its a no issue really.

My own experience since launch is that they behave very different to any other cpu, may it be clocks, temps, power consumption.. You name it all!
Undervolt and it'll say same frequency but performance lowers, temps are always high jut load doesn't really increase it much.
Idle has hardly any power draw.

 

[MrPotatoHead]

You know it's a good day at TPU when the only complaint about AMD is the BIOS number is confusing compared to 1st and 2nd gen Ryzen which although made great strides into gaining marketshare back from Intel wasn't quite good enough, now we have the same or greater IPC in instances I guess people have to bitch about something...

 

[Midland Dog]

Why not call it _____.4? Any f*ck having an in depth convo I hate typing on a phone be back in 2 days

 

[sutyi]

You know it's a good day at TPU when the only complaint about AMD is the BIOS number is confusing compared to 1st and 2nd gen Ryzen which although made great strides into gaining marketshare back from Intel wasn't quite good enough, now we have the same or greater IPC in instances I guess people have to bitch about something...

As much as I like AMD (and sometimes hate) people need to be critical when it is due. Some of these issues should've been fixed before the mainstream release.

 

[shmuck]

Slightly off-topic, but the new B450 Tomahawk BIOS (also released yesterday) allowed me to successfully cold boot this morning on the first attempt. I haven't been able to do this the last couple of days.

 

[the54thvoid]

Should the chipset driver be updated prior to a BIOS flash for the new 3rd gen? I'm on a CH6 at the moment.

 

[sutyi]

Should the chipset driver be updated prior to a BIOS flash for the new 3rd gen? I'm on a CH6 at the moment.

Probably does not matter. Might even want to reinstall the new chipset drivers anyway after swapping to 3rd Gen, cause it might install a different PowerPlan and such for a 3rd Gen CPU.

 

[CLevere]

The Data loss is not limited to third gen Ryzen, I have been dealing with that same issue using a 2600x and an NVMe drive. At least I am not alone, its been driving me crazy trying to solve that problem.