Linus Torvalds Is Sick Of AMD's fTPM Nonsense

For the past year Linux has been dealing with on and off issues with AMD's awful fTPM implementation and Linus Torvalds is finally sick of and it's completely being disabled from the kernel.
==========Support The Channel==========
► $100 Linode Credit: brodierobertso...
► Patreon: brodierobertso...
► Paypal: brodierobertso...
► Liberapay: brodierobertso...
► Amazon USA: brodierobertso...
==========Resources==========
Phoronix Article: www.phoronix.c...
AMD Intermittent Stutter: www.amd.com/en...
Kernel 6.1 Issue: bugzilla.kerne...
New Kernel Issue: bugzilla.kerne...
Asus Linux Website: asus-linux.org/
Torvalds Rant: lore.kernel.or...
=========Video Platforms==========
🎥 Odysee: brodierobertso...
🎥 Podcast: techovertea.xy...
🎮 Gaming: brodierobertso...
==========Social Media==========
🎤 Discord: brodierobertso...
🎤 Matrix Space: brodierobertso...
🐦 Twitter: brodierobertso...
🌐 Mastodon: brodierobertso...
🖥️ GitHub: brodierobertso...
==========Credits==========
🎨 Channel Art:
Profile Picture:
/ supercozman_draws
#Linux #AMD #LinusTorvalds #TPM #OpenSource #FOSS
🎵 Ending music
Track: Debris & Jonth - Game Time [NCS Release]
Music provided by NoCopyrightSounds.
Watch: • Debris & Jonth - Game ...
Free Download / Stream: ncs.io/GameTime
DISCLOSURE: Wherever possible I use referral links, which means if you click one of the links in this video or description and make a purchase I may receive a small commission or other compensation.

[@Turalcar]

Brodie: Linus Torvalds Subtitles: -12 volts

[@Wampa842]

Video: (makes literally any mechanical noise) Sub: [applause]

[@MateuLeGrillepain]

Subs when anything: foreign

[@DanielClear2]

from now on i will call Linus -12 volts

[@arnabbiswasalsodeep]

Oh my, subtitles for the 1st time have given me dad joke comedy gold

[@doigt6590]

Maybe that's Linus Torvalds real name and we just didn't know all this time!

[@mini_bomba]

I would've never suspected fTPM to be the cause of random audio cracks. What a weird world of proprietary firmware we live in.

[@iglobrothers645]

I actually figured it out pretty early after disabling fTPM and it stopped happening

[@Mallchad]

damn. and I was just complaining about TPM for the control and usage implications. Buggy syscalls are just NOT OKAY

[@iglobrothers645]

@@Mallchad yup especially when I know people who play valorant but can't anymore because either they get stutters or the can't play

[@DudeSoWin]

The audio is always first to get hacked to wiretap.

[@nasenbaer4627]

Ever since I've first started using my current Zen 2 based PC, I noticed audio would sometimes stutter/crackle. "Sometimes" meaning once or twice per day, for at most a second. It's mildly annoying, but not annoying enough for me to have ever looked into it. Until now I've just shrugged it off as a byproduct of a somewhat unstable OC or something being wrong with the soundcard(-portion of my particular mainboard). I would've never suspected the (f)TPM either.

[@igordasunddas3377]

"Nobody knows what's happening - except for the developers" - don't assume, that just because someone, usually a team, developed something, they really know how and if it works. I know from experience, that often developers don't know either.

[@BrodieRobertson]

They at least have a chance to know, whether they do or not that's fair

[@Bob-of-Zoid]

He is speaking of AMD's developers, and it's an AMD system they developed, so yes they should know, unless they don't know what they are doing in the first place, to which they shouldn't have released it until they had all the bugs worked out!

[@spl420]

​@@Bob-of-Zoidyou really underestimate "i made it work and have no idea how" factor.

[@afelias]

As a software QA I feel this way too much And as the QA that sometimes talk to the Support Team I also feel the other side of not being able to answer that kind of question completely

[@CraftMine1000]

Developer here, can confirm

[@davidjohnston4240]

RNG designer here : The idea of a firmware TPM having a secure RNG is just silly. A secure RNG needs a nondeterministic component, meaning it needs a hardware component.

[@Sandeepan]

All you ever need is a PRBS and a bit of love 🙃

[@CallousCoder]

A nice piece of radioactive cesium will be great to make a good but slow RNG. Or even using it just to reseed a software RNG frequently, making those a bit better. All other solutions will have a very pronounced Caussian spread even noise generators - although unique to each implementation. Now noise generators especially open collector are a terrible way to make random numbers (although very fast), but it’s easy to spoof with a transmitter sending out a predictable pattern pushing the odds in your favor. But probably knew that as an RNG designer - who had infinitely more experience with that than me. I only played around with some radioactive sources and detectors for a few weeks to get a very low P value random number generator as a learning thing. And even then the detector added a caussian spread, small but still noticeable.

[@davidjohnston4240]

@@CallousCoder In Intel chips, a latch is forced into metastability using a feedback loop which then generates 2.6Gbps of random data. This goes though a cryptographic entropy extractor and out comes fully uniform random bits. This uses a tiny bit of silicon area. The entropy comes from the thermal noise in the gates of the transistors in the latch. It is far far more reliable and faster than a radioactive source and has none of those pesky radioactive materials to deal with.

[@CallousCoder]

@@davidjohnston4240 the radioactive material isn’t the problem, people don’t realize we are bombarded continuously with charged particles and we are actually pretty resilient to them. The biggest issue is lack of speed in these quantum RNGS - unless you decrease the half life time then it gets increasingly more and more dangerous , and the problem is that your RNG runs out quickly too. I used Polonium-210 for a bit during my tests to get more randomization for the P tests. But with a half life of only 140 days it means that after 6 months you have no random numbers appear at all 🙂. Thanks for the deep dive in intel’s RNG. Yeah the crux is to reduce semi conductor material and avoid external interference. I didn’t know they did it like this. Cool! Is there any documentation about “entropy Extractors” I’m curious as to how that works.

[@ukyoize]

Can't you use graphic card to catch solar radiation?

[@C1TRU5]

Oh my god, I was going crazy trying to figure out what was going on with my PC. Thank you for posting this, this was exactly what I was experiencing

[@AutismusMaximus]

Same thing here Months passed and i could not figure out, what was wrong I am really happy now

[@absalomdraconis]

Good news: the AI that rootkitted your computer has a crush on you. Bad news: the AI that rootkitted your computer think it's funny when you get confused.

[@kofiy]

I'm using windows, but I still found out about that problem with fTPM only now...

[@CallousCoder]

You gotta love Linus’ pragmatism.

[@david-stephenson]

These stutters have been driving me crazy. I thought they were an issue from changing my GPU from Nvidia to Intel Arc and Nvidia drivers wrecking some configs. I had no idea it was because of my processor. It makes sense that I get these stutters because I had to upgrade to kernel 6+ to get the Arc card working. 👍Thanks AMD, I almost reinstalled Fedora hoping my Arc drivers would fix itself.

[@remasteredzero4076]

Fedora is amazing m8

5:17 Re Device Naming: The device is named in this "weird" way because of regional variants and general variant such as different hardware or revisions. In 99% of the cases it does not matter but sometimes it does when for example a older revision has a hardware bug, you need a specific device for example for having a foreign keyboard layout (e.g. ANSI US) or when different versions feature parts from different vendors such as for the screen or ram.

[@VincentVonDudler]

I didn't know this was an issue. But I have run into several other threads where TPM became an issue on Linux so I just disabled when i fully migrated over from Windows. It sounded like a lot of trouble.

[@looncraz]

Yeah, I just disable ALL TPM/security hardware. If I need security hardware I will attach one with USB... that's why I have a Google Coral :p

[@TremereTT]

@@looncraz Also there isn't much security in a PC that is USB comaptible anyways.

[@thechroniclesofcriss942]

I don't update my BIOS because there's always that warning that doing so can brick my computer, and I just go "nah, f--k that".

[@raiyankhanmajlis6112]

It took only a couple of seconds to do on both my low-end h61 and high-end z77 motherboard. That being said, both these motherboards are from 10 years ago (intel 3rd gen) so IDK about modern motherboards.

[@erk_0483]

it onloy briks your PC if it loses power during the installation, so the risk is not very big.

[@flarebear5346]

​@@erk_0483Really depends on where you live. Even in Lisbon portugal I still lose power randomly, I personaly only risk it when I have access to a ups

[@bleack8701]

​@@erk_0483I literally lost power today

[@Treznor.]

@@erk_0483 But its still there. I guess UPS is a thing that exists and would help with that risk a whole lot, but I'll guess most probably don't own one

[@raypol1]

Yeah had that issue when I bought my ryzen system. I though it was memory or something else until I found out it is related to TPM and disabled it straight away. Never had any issues after that and I do not really care about it anyway.

[@kilobytecache6192]

I've always disabled tpm. Never noticed anything stop working. I vote for simplicity

[@fmo94jos8v3]

you and I are alike at least in that regard :)

[@Xenotypic]

same

[@ThePortuguesePlayer]

I don't even know what TPM is. I don't think my CPU even has it.

[@Xenotypic]

@@ThePortuguesePlayer trusted platform module, came out around the windows 7 or 8 era if I remember right. It was touted as a security feature

[@ThePortuguesePlayer]

@@Xenotypic A random number generator is a security feature?... I'll probably have to read up on it.

[@bluesillybeard]

I was having stuttering problems a few months ago that seem to have mysteriously vanished. This might be related, although it could also be other things like kernel changes, driver updates, compositor updates, GNOME updates.... Really, it could have been anything, I'm just glad my system runs smoothly now (:

[@Lorentzeus]

You just fixed this annoying issue that I didn't even know how it was happening, such an easy fix for a month's worth of suffering. Thanks a lot.

[@arjix8738]

who do you refer to when you say "you"?

[@miaorenfeng3620]

just what I always wanted, a proprietary enclave in my system and it is buggy just for the kicks

[@squillium7149]

I’m glad you made this video this has been happening with my Linux machine a couple times a day and it was on my list of things to look into. I’m glad they have done work to fix this for Ryzen cpus

[@iguanac6466]

TPM...something consumers never wanted in the first place.

[@bes12000]

I was encountering this issue, I made a custom windows 11 install that removed the tpm requirement and then I went in bios and disabled it and reinstalled windows, no issues now, and my linux install works fine also, since im dual booting. And that BS line from microsoft about no updates if you install Windows on a system that doesn't have the TPM requirement, yeah my windows 11 has big updates all the time, no issues.

I think an issue existing inside the firmware can happen to any vendor. Updating firmware in a way like it does with ucode would be an idea however I didn't think it will work this way with something "secure" as a TPM where signatures have to be checked, the UEFI and the downstream motherboard vendor is involved. More vendors need to support BIOS/UEFI Updates with Windows Update/LVFS -> UEFI Capsules.

[@ericneo2]

Problem is they all cheap out and get tiny ROM chips and sooner or later run out of space.

[@davocc2405]

Wow, hadn't heard of a special project to optimise Linux for Asus kit - that actually doesn't surprise me, my experience with Asus as a company over the past 25+ years hasn't been.... well... ideal. They displayed an awful tendency to abandon kit that wasn't that old, even really high end stuff (speaking Windows here but the support culture was sub-par). that coupled with a really high failure rate - at one point I'd had 20 Asus based systems and only two hadn't failed horribly - I rank them as a brand to avoid, I ignore their fanboiz.

[@HappySlappyFace]

Are we talking entire systems or parts? As someone who uses Asus ( motherboard and used to have an asus gpu ) I haven't had any problems, the opposite honestly.

[@davocc2405]

​@@HappySlappyFacemixture of components, peripherals, entire systems, OEM builds, the works. Even monitors. I have odd exception bits which are ok - their original netbook EEEPC 1000H was one of the best in genre (I maintain the netbook era was highly underrated, I still have and use one actually). I had 2 out of 20 board not fail - had major problems like a top-end board with SATA connectors falling off; abandonment of $1000 video cards which needed ASUS support as they had all sorts of TV capture hardware; screens which *HIDEOUS* control circuitry and high pixel dropout rates. Oddly their networking gear has been really good that I've found so far, their routers are good enough t make me think they're outsourced. Quite a contrast to the rest. But overall - GPUs, mobos, controller cards, etc. always give me the expectation of near future failure. I've been a little more impressed with some of their bare bones systems which weren't *too* bad, they probably had corporate clients breathing down their throats about those. Over the decades I'd estimate an average of around 80% for serious problem/failure of their gear, driver abandonment was basically universal after about 18 months beyond model expiration. To contrast - saw a Gigabyte board from early 2019 that was receiving significant BIOS updates in July of this year, while yes it probably really needed it they were at least doing it.

[@extended_e]

I bought newly released ASUS computeronchip And on relase it already had all drivers outdated And only supported custom version of debian build at ASUS. That Debian version had been deprecated 5 years ago. So agree Asus got on ban list on my booms after that too. A fresh on market has 5 deprecated software on it and hasn't updated since.

[@cameronbosch1213]

I had the same thing on my Legion 7 2022 (Ryzen 7 6800H & Radeon RX 6700). I don’t know if Lenovo has the option in the UEFI but now I know why this happened.

[@Пердорыг]

So much time has passed... I kept thinking, what's the problem? And only now, two years later, I was able to solve the problem of freezing in games and the system. I thought the reason was a faulty SSD, but no, it was completely unexpected. I found a tip on the website, and then looked at the gameplay of one race with exactly the same problem. I went into the bios and turned off this bullshit.

[@andljoy]

TPMs , totally fantastic idea. Always work never cause problems!

[@autohmae]

Microsoft made a great decision to depend on them too...

[@Bob-of-Zoid]

@@autohmae Microsoft also made the terrible decision to want to be the sole arbiter and authority of the technology, and make everyone have to go to them in order to use it! And of course they either paid off or strong armed Motherboard manufacturers to put them on their boards, and tell owners to go to M$ for support even if you want nothing to do with them! Microsoft is a huge poisonous leech! It's why AMD came out with fTMP in the first place, and it's also why using it on Linux, BSD, Haiku and other OS's is so unwieldy! Microsoft came up with the tech in the first place too and holds patents on it! It's the kind of thing that should be open source and completely up to the user to use and control it, without having to go to some authority who anoints themselves co-owner of your computer!

[@vram1974]

Pluton coming to the rescue!

[@steeviebops]

@@Bob-of-Zoid You also need to remember what their original plan for TPM was. Look up Palladium or NGSCB to see some of the sneaky DRM plans they had. It's why I never 'trusted' the TPM (pun not intended) and Windows 11's insistence on having one means I haven't adopted it on my personal devices and maybe never will.

[@autohmae]

@@Bob-of-Zoid I think you are confused about secure boot/UEFI, which is something else.

[@RichardBetel]

JEEZUZ. I'm on an AMD, and occasionally see stutters. I need to look into this.

[@hyoryo]

yea, when i turned to the forums because i could not figure out why i suddenly got those stutters, i quickly learned about fTPM. since i don't use tpm at all, i simply deactivated it in the bios. a few months later a bios upgrade was supposed to fix this. but, the stutters got even worse. so, tpm is off again. and the "best" part is: those stutters are completely random. even by oversaturating the tpm with requests you can not trigger this!

[@nathanl2966]

I have been having this problem, figured it was the 7000 series needing more time to cook but apparently this has been around for a long time. Hopefully it is fixed and it allows me to safely crank my RAM to 7500MT/s as a assume these stutters are the reason it is unstable when I do so right now. 🤞

[@tcesportovo4284]

I have a R7 7730U is it going to get fixed just updating

[@nathanl2966]

@@tcesportovo4284 Yes, it's fixed since kernel 6.5. No more stutter.

[@hirose3871]

Ah, so that's what that was. I thought I had already disabled TPM but I certainly will now.

[@dunklerKurfurstDesDeepstat]

fuck! THANK YOU, that explains it! Oh got no more stutter, I just disabled that shit and now ... WOW, you made me so happy.

[@kabalder]

The funny part is that the stuttering issues now are caused by the windows spp-service using a really weird "let's place a phoney service far in the future at some point, because when this service can be changed, this means the spp-service is current and has the abilities it's supposed to have"-method. And it causes problems, because the component ignores the processor states of the cores when it's making the request. Meaning that you could get requests that would be on a system-level of surpreme importance, but that won't be scheduled to run on resources that are available, be that idle cores or unavailable external storage. If your laptop is asleep, for example, this service can fail and then trigger tpm measures like a reboot. Because, basically, nothing is as important to a properly running secure boot than checking the stored license-key every four and a random seconds. This is probably not made less problematic with the ftpm approach(the "f" just being that the "hardware" key is not stored on the platform's hardware - something Linus should welcome, frankly). But the "fix" that was actually rolled out wasn't to disable ftpm, or make sure the functions using them are not as exclusive, but to rotate response on all cores continously. There's a different pa (pa-400) that describes the issue from a "performance" point of view. And that's what is mixed in here. The issue being that the "fix" is basically to increase the amount of ambient burn on a processor by a large number of watts, with various other drawbacks. Which is not great on desktop, but specially bad on laptops. Btw, that something doesn't work on an Asus laptop is not proof that AMD's firmware model is broken. It's proof that Asus' outdated tool, with their weird inserts and hamfisted workarounds, is causing a problem. In the same way, I don't understand how a desktop that you didn't get from some OEM that insists on fusing a firmware module to it with helpful settings, etc... like enabling bitlocker, and having the firmware layer run weird programs to close out alternative boots, and enable "recovery" and so on.. would have that problem. I.e., you have to have been calling functions through the firmware that would have some other function in it. In other words, it's eminently possible that what's really causing the problem is some haphazard collection of commands on the firmware layer that always launches whenever the hwrng references are used. Testing also shows that this is not consistent behaviour even between the same hardware on different models from the same OEM. So this is not just haha AMD puts their tpm-modules (that we clearly need, right..) on external storage instead of multiplying the cost to get an Intel "trusted brand" function in there that clearly also always works, etc. It's something else. And the various workarounds triggered by this bs, to address the potential problems.. that most people don't have, after all. Given that you stay clear of Asus, for one.. or these OEMs that fit their security modules with clever programming, with assumptions that adding animated buttons and file refreshes in firmware functions can't possible go badly. -- these workarounds that are supposed to mitigate these stutter-problems are causing huge problems for the rest of us.

[@Skeleton-wn2zu]

Somewhat unrelated to the video's content but I like how you write new things on the whiteboard every video.

[@gamingonlinux]

Jeezus feck, I've been wondering for *ages* what's been causing my system to stutter constantly. I could be writing an article and suddenly the text stops being entered and then a few secs later catches up. SO FRUSTRATING. Upgraded BIOS today as apparently there was some fix from ASUS a good while ago too...

[@BrodieRobertson]

We love AMD

[@LINUXPENGUIN]

Ow, I have this issue on 6000 Ryzen CPU from moment I bought my PC. I have updated BIOS adressing this issue. Stutters are random and disabling fTPM completely - is the only way.

[@xmvziron]

I'm pretty sure I was affected by this, since I own a Ryzen CPU...

[@anterprites]

Interesting, I started having random stutters while playing VR games. (AMD 5800x3D and nvidia 3080) which I could not pin down to anything specific. But I don't use fTPM on Windows 10 (or it is used because I have it).

[@iannorton5514]

Holy crap i've been experiencing this for months! So glad i'm not going insane. Ryzen 7 1700x on an asus B350 mobo.

[@nikos4677]

I was having problem with tpm a few days ago after fresh install of arxh linux. I couldnt find any fix other than just masking the problem litteraly and metaphorically

[@Takyodor2]

Wait, did this cause my audio crackling sometimes, until I updated my BIOS + kernel a while ago?! 😮 I learnt something today!

[@BluuSkyz]

I've had issues with audio stutters, mouse stuttering and game stuttering for a while now and it took me so long to find out it was this.. 2022 Zephyrus g14. Driving me crazy

[@kyzo_124]

I had this for like a year and never knew what the problem was.

[@SkyString_]

On my Asrock B550 Taichi I was having this problem with Windows 11... After a lot of support tickets and BIOS versions "patches", the only fix was to buy a dedicated TPM module (TPM-SPI) and the issue never got back... Crazy this is going on years and they dont fix it... That and PBO2 values (TDP, TDC, EDC) more specific EDC, that in previous versions you could set above 140A and no performance lost was observed, and now anything above 140A will drop your performance a ton... These magic changes without a changelog or justification to a customer is the motive that in a next build I will go with Intel...

[@kleinstarnull]

Man....you've gotta be kidding me. I switched off of Arch to Windows because the stuttering (Ryzen 7 7735HS, Rembrandt) was unbearable. I kept seeing articles saying it was fixed, and gave up. I don't think it's completely 100% gone on Windows, but I can actually use this PC. Maybe with this, I might be able to switch back.

[@BrodieRobertson]

If that's not the issue now that's one less thing to worry about

[@unfa00]

Wow. I finally know what was going on when I used Ryzen 7 1700. I use 3800X now, and it doesn't happen, but darn...

[@TechnologyGeek862]

Well it seems that I’ve found a reason to my random stutter 😂 only thought it was my problem. haven’t had time to deal with it since it only happens quite rarely. 😅 might as well be fixed on my pc since it has been little while last I’ve seen it happen

[@BrodieRobertson]

Give it a try at least, if that's not the problem that's one thing you're sure about when looking for a solution

[@gutter_onion7855]

OMG... This is what I was dealing with for a while on my AMD Linux Desktop, and TRIED my hardest to narrow it down, I was going down the video/audio troubleshooting route, because I noticed it happened watching videos the most.

[@vedranb87]

Oh ffs... I finally know wtf was wrong with my AMD based Omen. No amount of research into audio stutter problem ever resolved the issue... And HP's bloatware never thought to prompt me to update BIOS. Eventually I replaced pretty much everything except the RAM and CPU and with the new MB from ASUS everything was fine from the get go and I wrote it off as a motherboard issue. Disabling TPM, apart from never crossing my mind as a possible cause, was never an option as I am required to use BitLocker as per company policies.

[@malmiteria]

i've had soooo many of those stutters for so long now. Playing trackmania, a *precision* game became impossible, i couldn't play for more than a year now. imma fix that shit immediatly

[@JorgetePanete]

It almost seems like propietary software and hardware should be considered unacceptable... I was in a job interview for traffic control in Java and the interviewer said they didn't use ANY open source programs, but they trust whatever crud they use

[@lixou]

:0 wow I’ve been having issues on windows for a while now (also AMD) where it just gets the whole screen ugly dark green for a minute or so and then goes normal. Maybe that’s also because of fTPM? Is there a way to fix it, but still having the TPM enabled? Maybe just a mobo firmware update?

[@fgsaramago]

I experienced this problem. First on windows 10, random freezes, manifested as stutters in games. This was in 2021! its what made me switch to linux on my main machine, thinking I would solve the problem, of course it didnt. Only i n May this year did it finally go away after I disabled all the tpm related crap. During this period I never found any info online that couldve pointed me in the right direction. I just tried different bios settings, linux distros and desktop environments.

[@bitterseeds]

Honestly, I forgot that I disabled fTPM before I even installed Linux on the new work rig and had to think for a sec ... nope no stutters. :D

[@EvanOfTheDarkness]

I've disabled TMP the moment I turned on my system. It's *just* a hardware DRM, that Windows and other software use to fingerprint and lock down your system. You are better off without it.

[@krashd]

You are better off without it only if you are computer savvy and don't install fruity apps, the millions of zombie botnet PCs around the world causing billions in ransomware damage and jeopardising lives shows that not all PC owners are computer savvy.

[@andreivaughn1468]

hardware TPM is only if you're ever worried someone will remove the hard drive from your laptop, steal just the hard drive, and try to read data off it externally. Happens really often, right??

[@cinemaipswich4636]

I am glad that I am waiting till the last minute to buy my next workstation. I generally use Supermicro WS motherboards and have been scanning the channels when it come to Windows 11 Pro. I need more that 32 PCIe lanes (audio, video capture cards, sync card) so its Xeon or Threadripper for me.

[@l3xitscomplicated191]

I had that stuttering and i noticed it in Valorant, it was so frustrating, i was looking around for the reason even changed out my ram...

[@MrBlazingninja]

5600 3060ti 16gb ram I get event error id 14 TPM and Valorant is the only game that stutters

[@terryforsythe8083]

I had a problem with it randomly knocking out my WiFi. I would have to reboot to get working again. It seems to be better after disabling TPM.

[@JamesMCrutchley]

I'm running windows on my AMD 5950x. I have an Asus motherboard. I upgraded the bios repeatedly till about 9 months ago when this issue finally just went away. At least for my config. But I have since found out that I cannot disable TPM at all anymore. My only option is to buy an external module and plug it into the TPM 2.0 header on motherboard if issue reoccurs. I cannot downgrade the bios from current version. It is locked. Asus says this is for security purposes and will not allow users to downgrade. So now I am stuck if I have any issues as Asus does not sell the TPM 2.0 module and buying it off of Amazon seems like a truly desperate idea. I am in luck though as a local pc retailer decided to buy as many of said tpm modules as he could order directly from Asus and is now selling them at no markup. So I could spend 15 dollars and just buy the module but I honestly am beyond frustrated with both the motherboard vendor and AMD at this point. Why disable downgrade of bios? Why make it impossible to actually fix the issue by completely disabling the tpm module. I had a post deleted by AMD for hacking on forum when I mentioned the idea of downgrading with instructions on how to disable in older versions of firmware. I was sent a message saying, and I am paraphrasing, that it was completely inappropriate and a violation of some US law to break drm if you downgraded by bricking bios and using flashback. It was and is intended for recovery and using it in that manner was hacking and also criminal behavior.

[@Todd_Manus]

The fTPM stutter happened while I was watching this video. I am on Windows 11, and an AMD Threadripper 3970x, ASRock Taichi MB. When I am on Linux, I do not see the issue... since I am on a lower kernel. Good to know, I am not alone... I can take solice in that. I worked with an AMD engineer, but he wasn't much help. He had me jump thru a lot of hoops to get log file, dumps, etc.. and then just stopped returning my emails.... hahaha.

[@zayn7lie]

I have encountered and bared this issue for nearly a year. The stutter even become more seriously in Wayland than X11. There is only short stuck in X11 but in Wayland, the whole system is entirely freeze with repeatedly audio in last several seconds.

[@aDumbHorse]

So THAT is the reason why my pc stutters randomly. This whole time I thought it was just some windows update causing havoc. I might disable fTPM if that's the case and maybe switch to Ubuntu here too.

[@Wilker_uwu]

i've been having stuttering issues and just always assumed it was something about ram usage or swap memory being badly used. i might check this out sometime. ty for the news

[@gpisic]

Geeez thank you Brodie, you don't know how long i searched for my audio crackling problem, this finally fixed it. Damn you AMD, do things right or don't do them at all ffs.

[@thedoctor5478]

This has been a problem for a few years now, and it's still not fixed. The only solution on an asus laptop was to completely gut/disable TPM in the OS. It was patched in Linux kernel, and then a later kernel update brought back the bug, but even worse now. It got so bad that I had to install windows (Manually deleting the TPM driver fixes it), which is heresy for me. Maybe I'll try this custom compile flag.

[@recarsion]

Sooo I've been planning on getting a new PC and going full AMD, is this something I should be concerned about? Or just go Intel CPU + AMD GPU, that should work right?

[@eDoc2020]

With this patch you shouldn't need to worry about it.

[@snaj9989]

I had a Lenovo ideapad gaming laptop with h5600. I talked with multiple people that has the same laptop and the thing was getting BIOS updates regularly but none of them fixed the fTPM issue. So I just disabled it. Windows 11 can cope with it.

[@_sukuratchi]

Had the issue before and it drove me crazy. Updated my bios and was fixed on windows so I thought it was fixed. Can’t believe this is STILL an issue

[@dalriada842]

I have a mix of Linux, MacOS, and Windows systems. Two of my Windows machines have AMD processors. I've experienced stuttering in video and audio on both machines. I had no idea it might be down to fTPM. I assumed it was because Windows is a crappy OS! I did at one point think of getting the hardware TPM, as both motherboards have headers, but there was price gouging at the time. I may revisit it.

[@ABaumstumpf]

For a couple of years now we developers wanted to use AMD hardware (basically since the release of Ryzen). The IT department tested various configurations all the time but not a single time did we get a greenlight so far - cause every single time we had some major problems. Be it the complet breaking of security features, false-reporting of supported features, CPU lockup at lowest supported speed (gosh that was fun to figure out why a server randomly took over an HOUR to boot and then see the CPU at 400MHz), no longterm-support etc. We were still using some (ultra)lowpower intel laptops for development and the Ryzen 3xxx series laptops they tested had faster everything but could not be used cause they would not pass any of the security certifications...... still bitter about that but it is what it is.

[@ArveEriksson]

I... have a first gen Ryzen 5 CPU and don't notice any stuttering. Eh, maybe it just doesn't have TPM because it's too old, or something? I'm also rocking an Intel Core i7 870 on my Linux PC. Hey, it still compiles Chrome - you just need to give it a few days.

[@_risebeyond__]

I have a 6900hx from amd on my laptop it’s not Linux it’s windows 11 but I can confirm this laptop stutters 1-3 times a day randomly for 2-3 seconds. I have never found a consistent program that causes the stutter. I can confirm this issue was never fixed on most systems to this day

[@Ozzymand]

So this is why my PC has been micro-stuttering. It's been hoppening for months now and never knew what caused it

[@logannosleep5]

I've disabled the firmware TPM....updated to the latest bios... Made sure it wasnt temps. I searched and tried countless "fixes".... Still getting random micro stutters with a 5600x, and 3080 in lightweight games.... i can not wait to get away from AMD and windows, then never look back....

[@judewestburner]

Is this just for firmware TPM only? So if I'm using a hardware TPM we're good? How many AMD systems use fTPM? I've never ever heard of it before this video.

[@xmvziron]

Oh yeah, I heard about this.

[@jawbuster46]

I experienced this issue on ny system, I dove down a huge rabit hole diagnosing ram and even got a wrranty swap cou from AMD , i ended up buying a physicall TPM addon for my asus mobo, disabled ftpm and never had the issue again.

[@knghtbrd]

I didn't have this as far as I knew … except now I do? Static in the audio when watching videos, I'd been looking for a pipewire fix. Except it seems that pipewire might not be the problem! Problem found with 6.4.something on Debian sid, may have existed back in 6.3.something but I don't think so. Will be interesting to see what disabling the fTPM RNG does for me.

[@elzabethtatcher9570]

Here I am completely on Linus's side. I have a Lenovo Legion laptop with 6800H, and while I do not remember frequent stutters, it's daunting to think thta a buggy bios can fuck me up this way.

[@wildweasel3001]

It's calculating the NSA ECC backdoor 😂

[@ChrisSmith-rm6xl]

Now we are going to have to make you disapear.

[@adonespitogonaif]

I think I have this problem, I just didn't notice it that much since my workflow is not affected by it. But oftentimes I notice stutters when watching youtube videos on my amd 6800u laptop.

[@jbodden6977]

MINUS 12 VOLTS = LINUS TORVALDS (or vice versa)... NOW I KNOW WHAT CALIFORNIA COURT REPORTERS DO FOR SIDE JOBS...

[@DudeSoWin]

I had this issue too and it was crashing every damn thing imaginable.

[@Roxor128]

"This is the fun part about proprietary code: nobody knows what's happening except for the developers." And sometimes, not even them!

[@rancidbeef582]

As a software developer, I can say "usually" don't know.

[@Roxor128]

@@rancidbeef582 True. How many times have we found ourselves sitting around, scratching our heads and wondering where the hell did we screw up because our program isn't doing what we expected it to do?

[@AshnSilvercorp]

When we have TMP triggers at work. It's literally just scrapping whatever setup of the OS is on there and completely reinstalling. Why did it go off? IDK, some random install failure of a 1st party microsoft program. It failed successfully.

[@FireStormOOO_]

Ha, aren't you the optimist thinking proprietary devs know what's going on in their code. Keep secrets, don't keep documentation, and don't retain staff and it becomes a mystery pretty quick.

[@davidioanhedges]

Disable TPM completely seems like the perfect solution, it works, and the only downside is you can't use Microsoft Bitlocker, and Windows is a bit harder to install ? Both of these are wins

[@geoffreyvanpelt6147]

Who here is old enough to remember when MS used an html 3.2 rendering engine in a newer version of MS Office, instead of fixing the security bug in its html 4 engine? I wonder if something similar happened with the AMD firmware... someone grabbed an older file, made a tweak and called it good.

[@orbatos]

Almost like TPM was never designed to provide end user functions at all, and thus testing in house covers such narrow cases that it's useless. For receipts just look into how it was developed.

[@cryptic_daemon_]

I honestly thought these stutters that i been experiancing were just on my system, I didnt know just how widespread it was until now.

[@astralchan]

Amazing commentary ^^ Now I'm curious, does this mean they patched it so the default option is to disable it? If someone were to enable it, would it be likely they'd experience the bug, even now?

[@Marcosilva0000]

Just ordered a hardware TPM to put on a desktop windows system. I've been having so many problems with random stuttering, mostly noticeable with audio stuff but I guess it's just what's easy to pick up nowadays

[@curryleavesbydhanya]

I seem to have this same issue with AMD Ryzen 5500U. Never knew this was the cause!

[@djvidual8288]

Might this be the problem that Eposvox had? Because of random stutters he switched to Mac.

[@helebarda344]

This bug hard crashed my system so many times that I got sick of it. I had updated my BIOS and issue continued. Since I could not find a TPM module for my MB (b450 tomahawk) I switched my MB which finaly fixed the problem.

[@RitzyBusiness]

I still have issues with AMD fTPM on my windows machine despite my firmware and OS being updated. While the stutters are rare, they do happen and its highly annoying when they do. I installed a TPM chip to my mobo and that worked fine, but the cheap POS died on me pretty fast. edit: Just to add my Ryzen 5000 system has been a mess since I put it together. USB in particular was very bad, you'd have these dips and stutters that would result in multiple key presses. I had to install a PCI-E USB card to fix the problem. Its supposed to be fixed today by bios but I doubt it is fully since I still sometimes have intermittent USB trouble on the MB ports. I will probably go intel with my next build, both my core2duo and i7 builds had zero problems, and this 5800x build has been nothing but problems.

[@clockblower6414]

Delete windows

[@RitzyBusiness]

@@clockblower6414 Nope, I don't think I will.

[@krashd]

@@RitzyBusiness At the moment you can legally and easily modify Windows 11 to lose the TPM requirement and then just turn TPM off (lots of people have done this so they can install that OS), if you have Windows 10 then you can just turn off TPM without any modifications. If you are fairly computer savvy and not prone to having your PC taken over by malware then you'll be completely fine without TPM, like virtualisation it's just another hardware feature to make it harder to hijack a PC but the best defence against malware is always a user who is astute enough not to click on strange links or install strange apps.

[@Johnithinuioian]

@@RitzyBusiness Dual-boot Linux

[@charginginprogresss]

8:10 Wait, so now I may understand why on my HP laptop I keep getting the same error in the event viewer. It's an AMD key ID certificate error, it says it can't find it, when I googled it the other month I saw it was somewhat related to the tpm, though at the time I had no clue about anything else. It could also be related to this.

[@notgate2624]

Holy shit ive had stutters on my amd laptop for months and never knew what caused them

[@brei2670]

Had the stutter issue. BIOS update not available at the time and was looking to take at least another month or so. Bought a discrete TPM module for like 15 bucks or so and used that instead. No problems since.

[@blackmoon_pr4189]

My asus rog g513rm 2022 laptop with r7 6800h suffer from ftpm and asus refuse to give a bios update for it or solve ftpm problem and no one in youtube is talking about it thanks amd

[@Kneedragon1962]

Maybe I am missing something, but I thought that was all about Windows and their "Trusted Computing Module" without which the system can't (won't) run Windows 11. Ergo ~ if you dual-boot, then you need it, but if you run Linux only ~ and you can turn off or disable the thing in the BIOS, then ~ That said, if there was nothing in Linux that used the damn thing, then why start using it now?

[@AnotherLotte]

Yeah, this isn't the first time hwrng has caused issues for people. At the same time, I didn't think this one would be tied to fTPM considering that I've never really had an issue with any kind of stutters.

[@moetocafe]

I have issues trusting the hardware TPM in the first place. After all 3-letters agencies backdoors, seemingly implemented on our hardware on purpose, why would anyone trust the TPM at all?

[@CrackManT]

I'm just seeing an OEM & Linux issue here, not an AMD one, perhaps the OEM systems are using cruddy or outdated firmware for the fTPM PSP device, maybe it's their bios? Maybe it's the proprietary part of their bios aside from AGESA? If you can't change hardware settings in your laptop's bios or that's the only type of platform where this is happening and it's not even widespread or common among all laptops of the same maker, how can you tell its the fTPM implementation? There's also the possibility it's something about the way Linux accesses or interacts with the fTPM module within the CPU, being open-source does not mean its free of bugs. How come updated AGESA's under Windows don't have this issue anymore? What is Linux doing differently to the fTPM that has made this issue resurface? Windows 11 uses the fTPM heavily (if present) and I have not heard of a single stutter or issue at all. You know, I wish we could just start investigating things further before pointing fingers, its this kind of behavior that stains the open source community, mostly the Linux related ones because it's where it happens the most.