AMD AGESA V2 1.2.0.7 Microcode To Fix fTPM Stutters

Dazz

2022-06-15 01:05

Passus:

For my X570 Strix Gaming F the 1.2.0.7 was there as a beta since the end of April, now the bios is dated the end of May but are the same release just not marked as a beta now

Yeah I noticed that too as I have had mine since early May and it was marked as beta, now it's not but is the same revision.

#6025386

GamerNerves

2022-06-15 01:55

Since I updated to 1.2.0.7 on Asus TUF Gaming B550M-Plus Wifi I have had one very brief stuttery moment, much shorter than what occasionally happened on 1.2.0.6, and also an USB dropout. The latter is actually a more severe problem, since it cut out my microphone completely, because it reset voice settings of Discord when it occured - this is what likely happened. Not very convincing from AMD. I'm unfortunately starting to believe that this base issue of Ryzen CPUs, the occasionally occuring stuttering, might not be fixable, but perhaps it can be almost completely mitigated? This is just speculation of course, AMD knows the truth, but doesn't openly share it for some reason and wasn't honest when they claimed that it was about the fTPM thingy (or they just didn't fix the related issue completely?). What is certain, is that they are struggling in revising the microcode - perhaps it's just that much rework to do? They definitely should speed up this process no matter how much resources it will cost for the sake of their customers and brand credit. Remains to be seen if Zen 4 has similar problems. EDIT. clarification

#6025479

hawk7000

2022-06-15 12:05

rl66:

? I have this agesa since around a month or so...

Undying:

Same here but this one is dated 2022/05/12.

I think the confusion simply stems from this article referring to the fix in the future tense despite updates with AGESA 1.2.0.7 having been released for most boards some time ago. I think it's just a mistake in this article, the fix is already out for those boards that already have 1.2.0.7.

#6025530

AlmondMan

2022-06-15 15:33

cucaulay malkin:

where do you even put that tpm module ?

Your motherboard has a specific header for it if it supports discrete modules. I have one installed.

#6025566

cucaulay malkin

2022-06-15 17:53

AlmondMan:

Your motherboard has a specific header for it if it supports discrete modules. I have one installed.

you plug it and that's it ?

#6025585

blkspade

2022-06-15 19:05

cucaulay malkin:

you plug it and that's it ?

Well there are 2 different types with different pin counts, so you need be sure to get the one that fits your board. Then you have to enable the discrete TPM option in your BIOS.

#6025602

PapaJohn

2022-06-15 19:55

And still Asrock don't deliver for my shitty board, Shamrock more like

#6025603

Ener

2022-06-15 20:01

mature platform btw

#6025648

Astyanax

2022-06-15 23:37

Kaarme:

ROG CROSSHAIR VIII HERO manual says the TPM module is purchased separately, so it wasn't shipped with it. Of course you can always have a look at the mobo and see if the connector is empty, to be absolutely sure. I'm not sure if any modern mobo is shipped with dTPM since modern CPUs should be able to handle it (fTPM). Maybe some professional mobos do, like Threadripper ones. No idea. Based on what folks have posted in the past, with at least some mobos, it seems possible to select dTPM in BIOS even when there's no module installed, which has led some to think they have a dTPM module, despite not possessing it in reality. It's kind of strange. A TPM module costs about 15 euros, so whenever I'll upgrade to Win11 (not any time soon), I might be tempted to get a dTPM just to be sure and avoid any annoyances. Time will tell.

all ryzens have an ftpm, the issue only affects the ftpm and only where there are un-resolved pcie-whea errors

#6025654

Ener

2022-06-15 23:41

Astyanax:

all ryzens have an ftpm, the issue only affects the ftpm and only where there are un-resolved pcie-whea errors

I have experienced this fTPM stutter but there are never WHEA errors logged. There's only one in my event log but that was during a y-cruncher test from when i was trying to massage 1900 fclk into submission.

#6025659

Astyanax

2022-06-15 23:55

Ener:

I have experienced this fTPM stutter but there are never WHEA errors logged. There's only one in my event log but that was during a y-cruncher test from when i was trying to massage 1900 fclk into submission.

Thats because AMD disabled the error reporter for non-critical AER's.

#6025713

Kaarme

2022-06-16 09:09

Astyanax:

all ryzens have an ftpm, the issue only affects the ftpm and only where there are un-resolved pcie-whea errors

15 euros isn't such a huge amount that I wouldn't be willing to spend it to avoid trouble later. I did suffer from the usb disconnect issues before a bios update, so while I'm generally speaking happy with my Ryzen, I don't trust it 100%. Considering how awfully long this supposedly fTPM stutter fixing update took AMD to develop, I trust it even less. The longer something quite critical takes to mend, the more fundamental and possibly unfixable the underlying problems are. Just like the Spectre and Meltdown problems that plagued multiple generations of Intel CPUs, and still appeared to some degree in CPUs Intel released after those security holes were well known.

#6026067

AlmondMan

2022-06-17 13:22

cucaulay malkin:

you plug it and that's it ?

Yes. Then I ensured that it was set to discrete TPM in my BIOS settings. But like it was mentioned, you need to get one that matches. Prices also increased 100% around the time that Windows 11 announced it required them. For some reason. I'd no real need for it, I just enjoyed the notion of having another chip plugged in. I enjoy retro computers and all the ways you used to plug in connectivity via more boards, so this was mostly a fun thing for me to install.

#6026082

PrMinisterGR

2022-06-17 14:01

fTPM should be completely fine, especially after this minor fix. I had some random stutter like once per week perhaps, but nothing after 1.2.0.7. A TPM is just a place where random numbers can be stored securely. It's closer to a password manager, than a DRM module. It can be used for DRM, but then other things can be used too. If anyone is really interested in fighting DRM, the only place this can actually happen is on the legislative and nowhere else. I would prefer to have a TPM installed and just pirate GOG games, than having weak cryptography on the system I'm using.

#6026130