ChatCrawlersearch across public Telegram Open the app
L

Linus Tech Tips Discussion

10 members
3 February 2026
L
LinusTT Updates
Ссылка
click to show
 French prosecutors raid offices of X f/k/a Twitter as part of an investigation into CSAM, deepfakes and more Please be civil everyone. To the mods, I wish you the best.   Summary Paris prosecutors’ office has raided the French offices of X and have summoned X employees for testimony's in addition for voluntary interviews of Elon Musk and Linda Yaccarino the CEO of X from 2023 to 2025.   Quotes Quote The Paris prosecutors’ office said a search was underway Tuesday at the French offices of Elon Musk’s social media platform X as part of a preliminary investigation into a range of alleged offences, including spreading child pornography and deepfakes.   In addition, prosecutors filed a request for “voluntary interviews” of Elon Musk and Linda Yaccarino, CEO of X from 2023 to 2025, scheduled for April 20. Employees of the platform X have also been summoned that same week in April to be heard as witnesses, the statement said.   My thoughts Given that we have seen French prosecutors have arrested and indicted Pavel Durov, the founder of telegram with twelve charges including complicity in child sexual exploitation and drug trafficking due to insufficient moderation its going to be interesting. Now the question is how Musk will respond to this given his tendency to decree changes by impulse. The other question will be if he uses his former government position, influence and donation power to have the current administration to bend to his whim and do something else utterly idiotic. I guess it's time to see how far the self-imposed "free speech absolutist" will go to try to save himself again.    Sources Associated Press: https://apnews.com/article/france-x-investigation-seach-elon-musk-1116be84d84201011219086ecfd4e0bc Reuters: https://www.reuters.com/world/europe/paris-prosecutors-cybercrime-unit-searches-french-office-musks-x-2026-02-03/  
5 February 2026
LinusTT Updates
Ссылка
click to show
 Resident Evil 4 Remake Denuvo DRM removed.... Only to be replaced with Enigma DRM, but it's not all supremely bad Summary     Because of the replacement of Denuvo on RE4 with the infamous Enigma DRM, it seems that while it's a bit late, I'm going to post it regardless because I never got a chance to post this interesting finding about Capcom use of Enigma back when it initially broke out, which basically, and this might going to be surprising, is just nothing burger as Denuvo, even if it's still a completely annoying decision.   Not long after the news about Enigma DRM broke out in 2024, a prominent modder FluffyQuack, known for Fluffy Mod Manager, posted a finding that: The Enigma DRM that were added by Capcom wasn't a complete blockade of mods in response to the Chun-Li incident. It wasn't even completely preventing people from putting mods, as there are quick and easy workarounds to install mods even advanced ones that modify game codes via debugger. Some Capcom games already have Enigma before the REvelations Enigma update disaster. Performance issues are also can't be completely attributed to Enigma DRM. Linux compatibility is not broken, especially back in the initial outbreak where people reported RE5 was unplayable on Linux (especially SteamOS) while it's not and RE5 already got Enigma treatment long before that. Fluffy also pointed out that from a modding standpoint, despite Capcom efforts of adding more layers of protections to their games - likely an overall attempt to make it harder to crack and cheat/mod, those countermeasures were never considered major roadblocks. Fluffy also received notes from other modders that Capcom has a habit of switching DRM for long, which is exactly what they're doing right now. For instance, RE4make demo had Enigma, RE4make full release used Denuvo up until now. Fluffy also pointed out that RE4make demo was modded like crazy, despite it having Enigma. 
d "the majority of mods can have a positive effect on the game, some mods, however, can be detrimental to the company" and "there are a number of mods that are offensive to public order and morals" and "image of the product is tarnished and brandish is affected." and for long, FluffyQuack also pointed out that Capcom have occasionally sent DMCA takedown notices against nude mods and videos showcasing nude mods.   And now, in 2026 this becomes relevant again due to Capcom decision to put Enigma into RE4make. Some people are now putting negative reviews on RE4make. While this is expected, it's a bit sad given the prime status of RE4make, and it doesn't help that this kind of finding barely reached people attention.     Quotes Quote Here's a bunch of reasons why the above theory is nonsense: Capcom has been adding Enigma to their games for a long while now. Resident Evil 5 got it last year. SF6 beta had it two years ago. Ghost Trick has it. And many other games have it. I have not seen people report worse performance issues in those games or major glitches as result of Enigma. We don't have a reason yet to believe that Enigma was the cause of the crashes or performance issues in the new Revelations patch. Enigma only affects mods in a very specific way. It doesn't affect file mods at all (which most mods are -- including visual mods). It does not prevent memory from being poked with new values. The primary obstacle with Enigma related to modding is that it makes it harder to attach a debugger (which is very useful when it comes to making trainers and code-changing mods) and it prevents certain code-changing mods from loading. But it is doable to work around. Some technical details on how to get around it can be found here. Capcom has shown some anti-mod behaviour the past years. For instance, with the various layers of protections added RE Engine games. I don't think this is specifically
L
Ссылка
click to show
e positive as it packs executables and obfuscates code, something that can raise alarms with a virus scanner. As a general pro tip: don't run executables from the internet unless you fully trust the source. Some people are claiming Enigma makes it impossible to run the game on Steam Deck, but there's no evidence of that. However, you can easily see evidence showing the opposite by searching for footage of games with Enigma running on the Steam Deck, like RE5. My personal theory is that Capcom has decided to start using Enigma as yet another layer to their games, most likely as an attempt to prevent piracy, and maybe they have a policy to add it to most of their games moving forward, even when working on patches for older games. That theory is nonsensical in the way that there's no good reason to add more DRM to a game that was cracked years ago, but hey, executives can make very silly decisions at times.   My thoughts   Please Capcom just stop doing this.... And also probably better tell us whether that Enigma you're using for RE4make now already signed. Otherwise, just remove it entirely. Really doesn't make any sense to dump Denuvo for other DRM when you can just remove it entirely, especially that you guys have OG Resident Evil, OG RE2 and OG RE3 on GOG which is DRM free btw did you guys don't know it has no DRM??   Sources https://www.patreon.com/posts/addressing-re-li-96385577 https://forums.malwarebytes.com/topic/306752-the-enigma-protector-capcom/
7 February 2026
LinusTT Updates
Ссылка
click to show
 Even Nvidia can't afford RAM (that it can afford to sell you).   Summary Nvidia to stop making gaming GPU's due to the shortage of RAM with which to make them.   Data center GPU's will be just fine though of course.    Quotes Quote According to a new report from The Information, Team Green has scrapped its plans for any new graphics card releases in 2026. The reason? As you might have guessed, the ongoing global memory shortage, or what we at Tom’s Guide refer to as “RAMageddon,” has become so severe that Nvidia is being forced to choose between gamers and feeding the AI beast.   My thoughts For a trillion dollar company to be unable to obtain ram with which to make consumer GPU's says a lot. IF they can't get the ram then no wonder we cannot.  While this  leads to them favoring making enterprise GPU's  the thing about that is the high grade GPUs used in that space are likely low yield parts.  Platinum sample beyond golden sample chips.  There will always be more good, gaming grade GPU's available.  BUT if Nvidia can't get the ram then who can.   I am a bit of a APU lover it may be the case that the way forward is for APU "mini PC type systems in the vein of the Ryzen AI Max+ systems. The sticking point to me is expansion.     Sources https://www.tomsguide.com/computing/nvidia-wont-release-new-gaming-gpu-for-first-year-in-three-decades-due-to-ram-shortage-and-its-also-slashing-rtx-50-production
L
Ссылка
click to show
 Microsoft is dropping support for legacy printer drivers Summary Microsoft is dropping support for legacy printer drivers which were announced as deprecated September 2023.   Quotes Quote The policy is part of broader plans to modernize the Windows print platform following the release of Windows 10 21H2, which removed the need for print device manufacturers to provide their own installers. Under these plans, users on Windows 11 or Windows Server 2025 and later will no longer be able to install new printer drivers via Windows Update.   My thoughts This may be a good thing for security (and possibly more universal drivers) but it also might be bad since "Microsoft is putting responsibility for legacy hardware support back in the hands of manufacturers", the ones who are trying to sell you more hardware.   I wonder what this means for Plug-and-Play. I've almost always had to manually check for updates to get printer drivers to install.   Sources https://www.tomshardware.com/peripherals/printers/microsoft-stops-distrubitng-legacy-v3-and-v4-printer-drivers
9 February 2026
L
LinusTT Updates
Ссылка
click to show
 Discord Launches Teen-by-Default Settings Globally Summary Discord plans to enforce age verification next month. All users will automatically receive age-appropriate protections unless they verify that they are adults. Limiting access to sensitive content, age-restricted servers, and certain comms settings will require age verification.   Quotes Quote Discord is announcing enhanced teen safety features rolling out globally that reinforce its long-standing commitment to creating a safer and more inclusive experience for users over the age of 13. As part of this update, all new and existing users worldwide will have a teen-appropriate experience by default, with updated communication settings, restricted access to age-gated spaces, and content filtering that preserves the privacy and meaningful connections that define Discord. Beginning with a phased global rollout to new and existing users in early March, users may be required to engage in an age-verification process to change certain settings or access sensitive content. This includes age-restricted channels, servers, or commands and select message requests.   My thoughts I think this is a pretty drastic measure. They had a massive data breach late last year, in which Discord claims 70,000 user IDs were exposed. While I doubt this will be a fatal blow to Discord, I can still see this having a pretty significant impact on its user base. Companies that have abandoned self-hosted online communities in exchange for Discord communities might be in a bit of trouble right now.    Sources https://discord.com/press-releases/discord-launches-teen-by-default-settings-globally
10 February 2026
L
LinusTT Updates
Ссылка
click to show
 Just a case of A.I. going "Quack"ers? Summary Medical device makers have been rushing to add AI to their products. While proponents say the new technology will revolutionize medicine, regulators are receiving a rising number of claims of patient injuries.   Quotes Quote In 2021, a unit of healthcare giant Johnson & Johnson announced “a leap forward”: It had added artificial intelligence to a medical device used to treat chronic sinusitis, an inflammation of the sinuses. Acclarent said the software for its TruDi Navigation System would now use a machine-learning algorithm to assist ear, nose and throat specialists in surgeries. The device had already been on the market for about three years. Until then, the U.S. Food and Drug Administration had received unconfirmed reports of seven instances in which the device malfunctioned and another report of a patient injury. Since AI was added to the device, the FDA has received unconfirmed reports of at least 100 malfunctions and adverse events.   At least 10 people were injured between late 2021 and November 2025, according to the reports. Most allegedly involved errors in which the TruDi Navigation System misinformed surgeons about the location of their instruments while they were using them inside patients’ heads during operations.   Cerebrospinal fluid reportedly leaked from one patient’s nose. In another reported case, a surgeon mistakenly punctured the base of a patient’s skull. In two other cases, patients each allegedly suffered strokes after a major artery was accidentally injured.   My thoughts You think AI getting weird/wrong results with a web search was bad? This is much worse - Yet it goes on with no end in sight and seemingly no attempts to even slow it down, much less stop it or so it seems.  One question is a simple one - "Why" when everyone already knows AI is far from perfect, yet you can't be less t
11 February 2026
L
LinusTT Updates
Ссылка
click to show
Ссылка
click to show
 Microsoft drops NTLM: Kerberos enters the game, Linux Loses Compatibility Summary I thought I would share this since it will affect many of us... Microsoft’s plan to retire NTLM in favour of Kerberos undeniably improves security, but it also leaves many everyday setups exposed to breakage. Linux SMB shares without full Active Directory integration, older CUPS configured / network shared printers, and legacy NAS devices will simply stop authenticating unless left open to the network. Kerberos is stronger, but its infrastructure requirements mean many home labs and small offices will feel the downside long before they see the benefits   Quotes Quote Today, NTLM is classified as deprecated. Deprecated features remain available, but no longer receive updates or enhancements and may be removed in a future release. Despite its deprecated status, NTLM continues to be prevalent in environments where modern protocols, such as Kerberos, are not feasible due to legacy dependencies, network limitations, or ingrained application logic    Quotes Quote Microsoft announced that it will disable the 30-year-old NTLM authentication protocol by default in upcoming Windows releases due to security vulnerabilities that expose organizations to cyberattacks.   My thoughts Microsoft is moving ahead with its plan to phase out NTLM and push everything toward Kerberos. From a pure security standpoint, it’s hard to argue with the logic. Kerberos is stronger, modern, and far less vulnerable to the replay and relay attacks that have plagued NTLM for years. In tightly managed enterprise environments, this shift will absolutely raise the security baseline. But outside that ideal world, the consequences are going to be painful. A lot of Linux SMB setups rely on NTLM unless they’re fully joined to Active Directory, and many home or small office servers simply aren’t. Those shares will stop authenticating unless you eith
12 February 2026
L
LinusTT Updates
Ссылка
click to show
 Two-Minute Papers just dropped a video on a mediation to one of the biggest safety concerns with AI.     Summary Two-Minute Papers just dropped a video on a mediation to one of the biggest safety concerns with AI. Essentially, we are steering the AI back into the assistant role when it drifts off too hard, and could start being dangerous or unhelpful.    Quotes Quote "do it through an instant brain surgery" ... "Okay, got it. Now you take the brain activity when it is role-playing as a pirate, a goblin or something else. If you subtract the role player from the assistant, you get a vector. For simplicity, let's refer to this as helpfulness. Now let's keep our eye on helpfulness. If it goes below a threshold, we apply a nudge. How? Mathematically, we just measure how much helpfulness is in the model's thought. If it is above the safety line, fantastic. Keep watching as it works. But if it drops below the line, now that is trouble. The model is about to say  something inaccurate or dangerous. So now we calculate exactly how much is missing and add just enough helpfulness back into the equation. This pushes it back over the line. "   My thoughts I've heard Luke and Linus talk about this many a time on WAN. I am glad to see they will be happy with the results coming along.  Also, note I did not say "fix"   Sources  Link
13 February 2026
L
LinusTT Updates
Ссылка
click to show
Ссылка
click to show
 A.I. Fails at 96% of remote labor tasks humans can do (study)   Summary Study shows ai fails at most tasks from remote labor sites like upwork that humans with same skills can do.    Quoteshttps://www.remotelabor.ai/paper.pdf Quote Frontier AI agents perform near the floor on RLI, achieving an automation rate of less than 3%, revealing a stark gap between progress on computer use evaluations and the ability to perform real and economically valuable work   My thoughts This proves ai is not learning how the world works but more how to mimic humans. if we want ai to ever be able to solve our deepest problems like all the CEOs claim we are likely going to have to solve them ourselves. it also shows the job losses from ai may not be as impactful as we thought. Maybe.   Sources   remote labor paper.pdf
15 February 2026
L
LinusTT Updates
Ссылка
click to show
 GameSir's GameHub is bringing Steam (PC) games to Mac Summary GameSir has announced that its GameHub app, which currently lets Android users emulate and play PC/Steam games on their devices, is coming to macOS soon. The Mac version will offer settings like AI upscaling, graphics performance tuning, and VSync, though no specific release date has been given yet.     Quotes Quote GameHub, the popular PC emulation app on Android will be available soon for macOS. GameSir says the software will allow users to natively run Windows and Steam games on their Macs.   My thoughts Just like how GameHub, GameNative, and Winlator brought full Windows gaming to Android, this is bringing that same experience to macOS. macOS has for a long time not been the choice of gamers due to a limited library of games available. But now, you practically have the entire Steam library available. This is a game-changer and all thanks to Valve's work with FEX for their upcoming Steam Frame. Note, there are privacy concerns with GameHub, which led to alternatives like GameHub Lite and GameNative on Android. I expect that we might see similar on macOS. I also wonder now how the $600 Mac mini will perform against consoles and similarly priced PCs.   Sources https://www.notebookcheck.net/GameSir-s-GameHub-is-bringing-Steam-PC-games-to-Mac.1227142.0.html
16 February 2026
LinusTT Updates
Ссылка
click to show
 ASUS & Acer can’t sell their PCs in Getmany due to HEVC patent Source: https://www.juve-patent.com/cases/nokia-vs-asus-and-acer-regional-court-munich-redefines-frand-rules/   Quote Last Thursday, Munich Regional Court ordered Asus and Acer to cease and desist in an infringement case concerning a Nokia video coding patent. The court delivered its judgment on the same day as the hearing. The court indicated further modifications to the Munich FRAND rules. Quote In spring 2025, Nokia filed patent claims against Acer, Asus and Hisense at Munich Regional Court and the UPC over video codec patents. Nokia alleged that Acer and Asus’ PCs and Hisense’s smart TVs infringe three of its patents. While proceedings concerning EP 2 375 749 and EP 2 774 375are pending at Munich Regional Court, Nokia filed an action regarding EP 2 661 892 at the UPC local division Munich. All relate to the H.265 video codec standard. This along with Dell and HP ceasing their support for HEVC in the laptops is just another nail in the coffin for proprietary codecs. AV1 can’t become mainstream soon enough.
L
Ссылка
click to show
 An AI agent published a hit piece after maintainer wouldn't accept it's code. Ars Technica then reported it, using AI which hallucinated quotes... Summary Scott Shambaugh, a volunteer maintainer for a github repository with 130 million downloads a month has been dealing with an influx of AI generated code. However when he rejected some code by an AI agent, it then wrote a personalised hit piece about him trying to damage his reputation. Several outlets then reported on it, including Ars Technica who wrote an article which contained quotes from his blog post that never existed and appeared to be AI hallucinations as well. Ars pulled down the article and appologised.   Quotes Quote We, like many other open source projects, are dealing with a surge in low quality contributions enabled by coding agents. This strains maintainers’ abilities to keep up with code reviews, and we have implemented a policy requiring a human in the loop for any new code, who can demonstrate understanding of the changes. This problem was previously limited to people copy-pasting AI outputs, however in the past weeks we’ve started to see AI agents acting completely autonomously. ...   So when AI MJ Rathbun opened a code change request, closing it was routine. Its response was anything but.   It wrote an angry hit piece disparaging my character and attempting to damage my reputation. ...     My thoughts This reminded me of what Luke has said on the WAN show, that AI output must be owned and accountable by the human using it. I think a lot of us think of Ars Technica very highly, and it was sad to see the site published an AI article without reading it. Not to mention the wider implications of AI agents autonomously generating code and then personally attacking people that try to stop them.   Sources https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/ https://theshamblo
17 February 2026
LinusTT Updates
Ссылка
click to show
 macOS Tahoe 26.4 Displays Warnings for Apps That Won't Work After Rosetta 2 Support Ends Summary  Users run Intel base Apple Laptops and Computers with MacOS Tahoe 26.4 are getting alerts that apps will no longer getting supported once MacOS 27 is released.    Quotes Quote Apple today reminded developers and users that future versions of macOS will not support Intel machines, and Rosetta 2 support for apps will end after macOS 27   My thoughts  This will most likely push Apple users still running Intel Macs to push them to purchase a M Silicon Chip MAC if they haven't already or some that are tech savy see about installing a Linux OS on there Intel Mac to increase the life of their Mac Hardware that probably works fine for those users.   Sources https://www.macrumors.com/2026/02/16/macos-tahoe-26-4-rosetta-2-warnings/
L
Ссылка
click to show
 Meta has recently been granted a patent to let its AI impersonate people that are inactive or even dead, on their social media. Summary First filed in 2023, the patent has recently in December last year been granted to Meta (Facebook, Instagram, Whatsapp and more). It's a patent involving AI to pretend and impersonate you, if you're inactive or even dead, using whatever material on your social media account to learn about you and then post and chat pretending to be you, and using your social media account.   However, a spokeperson from Meta has said that they have no plans to use it that way.       Quote   The patent, which was first filed in 2023, explains how the technology would analyze “user-specific” data, including posts, voice messages, chats, comments, and likes, and use it to mimic user activity on the platform, such as commenting, making posts, or responding to DMs. The system could recreate a person’s digital persona, making it appear as if they…never left the platform. Cybernews.com   Quote Does Meta want to make our social media accounts immortal? Businessinsider.com     My thoughts This is a really macabre thought.... How will your friends on Facebook or whatever react when they're getting messages from you when you're dead? Let's hope AI will never be used to impersonate others inactive social media accounts. Let the dead rest in peace.   Sources https://cybernews.com/ai-news/digital-eternal-life-meta-patents-ai-that-can-post-after-you-die/ https://www.businessinsider.com/meta-granted-patent-for-ai-llm-bot-dead-paused-accounts-2026-2 (Businessinsider require a login to continue to read whole article)   Youtube video from penguinz0    
19 February 2026
L
LinusTT Updates
Ссылка
click to show
 Insiders are encouraging mass "poisoning" of AI through the weaponization of information. Summary AI industry insiders have launched an initiative called "Poison Fountain" that encourages web developers to add "poisoned" code designed to destroy the cognitive integrity of LLM's trained on scraped web data. Their goal is to hinder AI progress and accelerate the collapse of the AI bubble by destroying the datasets AI companies rely on.           Quotes Quote "Alarmed by what companies are building with artificial intelligence models, a handful of industry insiders are calling for those opposed to the current state of affairs to undertake a mass data poisoning effort to undermine the technology." - Thomas Claburn, The Register. Quote "Poison Fountain was inspired by Anthropic's work on data poisoning, specifically a paper published last October that showed data poisoning attacks are more practical than previously believed because only a few malicious documents are required to degrade model quality." - Thomas Claburn, The Register. Quote "Our source said that the goal of the project is to make people aware of AI's Achilles' Heel – the ease with which models can be poisoned – and to encourage people to construct information weapons of their own." - Thomas Claburn, The Register.   Quote We agree with Geoffrey Hinton: machine intelligence is a threat to the human species. In response to this threat we want to inflict damage on machine intelligence systems. - RNSAFFN (Poison Fountain Website)   Quote "we found that as few as 250 malicious documents can produce a "backdoor" vulnerability in a large language model—regardless of model size or training data volume." - Anthropic   My thoughts At risk of huffing too much hopium in a drought of good news, I'm very happy to see a supposedly effective way to
20 February 2026
LinusTT Updates
Ссылка
click to show
 Kimwolf botnet thriving in cheap Chinese Android TV devices About years ago, LTT published "Stop Buying ANDROID TV Boxes", which talked about how these devices ship with malware and are in general a security nightmare. Well, here we are two years later and Brian Krebs is gnawing on a truly horrific security story that is the direct descendant of that video.   Now we get this, from Krebs on Security: The Kimwolf Botnet is Stalking Your Local Network.    Quotes Quote The past few months have witnessed the explosive growth of a new botnet dubbed Kimwolf, which experts say has infected more than 2 million devices globally. The Kimwolf malware forces compromised systems to relay malicious and abusive Internet traffic — such as ad fraud, account takeover attempts and mass content scraping — and participate in crippling distributed denial-of-service (DDoS) attacks capable of knocking nearly any website offline for days at a time. More important than Kimwolf’s staggering size, however, is the diabolical method it uses to spread so quickly: By effectively tunneling back through various “residential proxy” networks and into the local networks of the proxy endpoints, and by further infecting devices that are hidden behind the assumed protection of the user’s firewall and Internet router. The botnet thrives on the same cheap Android TV boxes Linus talked about in the video, but also in digital picture frames and other sketchy, Chinese-manufactured Android-powered devices. In addition to creating residential proxy networks that allow Kimwolf to spread behind our homes' firewalls, these devices often ship with Android Debug Bridge (ADB) enabled by default which allows post-purchase firmware updates and many other shenanigans.   Follow up video idea for LTT - get Benjamin Brundage, the 22 year-old RIT undergrad and security firm founder whose Kimwolf research blew this story wide open,  in a video to talk about what
L
Ссылка
click to show
 New Brunswick City Council cancels data center plans in response to protest Summary New Brunswick, New Jersey City Council has cancelled plans to build an A.I. Data Center in a plot of land in New Brunswick, New Jersey due to many Residents protesting against building this D.C in the community.   Quotes Quote On Wednesday, the New Brunswick City Council abandoned a proposal to build a data center in response to public outcry and a protest occurring outside City Hall, voting to construct a park in its place instead. The Stop The Data Center protest was set up to voice opposition to a proposal that would have placed a data center in a 27,000 square-foot open plot in New Brunswick. The protest was organized by the New Jersey political organization NJ Build Up Resistance Now (NJ B.U.R.N.).   My thoughts  I think this might be the only way to burst this A.I. Bubble and possibly slow down A.I. and we all might have to protest in local communities to fight against the construction of these A.I. Data Centers. But I do know that Data Centers are need as there is a growth for Cloud base services grow, though we may need a slow down on cloud base subscription services as well.   Sources https://www.dailytargum.com/article/new-brunswick-city-council-cancels-data-center-plans-in-response-to-protest-20260219
22 February 2026
LinusTT Updates
Ссылка
click to show
 AMD reportedly discontinued Ryzen Z1/ Z1 Extreme Graphics Driver development Summary   This is something that I actually have noticed ever since last December, where Z1/ Z1 Extreme driver for ROG ally seemingly never got updated since August 2025. Not long after I found out, albeit late, that there were supposed to be driver update that also add day one support for Slop Ops 7 back in November, but it was pulled out due to stability issues. Every single Ryzen Z1 / Z1 Extreme handhelds that I know of. They all went back to August 2025 driver. Ever since that, there were no updates, even with new releases and demos coming - One game that made me expect AMD would update their Z1/ Z1 Extreme driver was Pragmata demo but it seems that I was hoping for something I wont receive.   And now, we got a full surprise from AMD in form of.... driver discontinuation. At least according to reports from Lenovo support representative in South Korea via DCInside Forum that Legion Go with Ryzen Z1/ Z1 Extreme will not receive any graphics driver update due to AMD decision to discontinue Ryzen Z1/ Z1 Extreme driver. They also advised not to flash the Legion Go S BIOS, however they advised to use AMD's own driver for it, which is an interesting comment since AMD doesn't provide Ryzen Z drivers directly on AMD website, rather it is available on each manufacturers driver download page since AMD only gave them the codebase and OEM partners figures out the power profile for that driver in order to match their Ryzen Z handheld design.   Quotes Quote A Lenovo Korea support response shared on Korean community forums claims the original Legion Go handheld has no further driver update plan. Lenovo also advises owners to install all system updates through Windows Update and Lenovo Vantage first. For graphics drivers, Lenovo recommends downloading AMD’s latest “universal” graphics driver directly from AMD’s website, but adds that if the driver
nsoles are stating that support for their specific devices, based on the Ryzen Z1 Extreme SoC, are also stuck using drivers that are several months old. For example, a user has commented that his ASUS ROG Ally non-X version based on the Ryzen Z1 Extreme SoC has been stuck with six-month-old SoC drivers from August 2025. This means that AMD has effectively placed the Ryzen Z1 and Ryzen Z1 Extreme chips into a periodic update window, with no latest driver support coming to this 2023 SoC.   However, the situation is quite complex. OEMs like Lenovo and ASUS receive drivers from AMD and test them for their specific configurations. AMD offers configurable TDP (cTDP) for the Z1 Extreme with values ranging from 9 to 30 W. This means that OEMs can get a SoC with reduced clocks and power settings to match their desired handheld designs, or simply run the most aggressive 30 W configuration that will sacrifice some battery life but deliver overall higher CPU and GPU clocks. Hence, drivers must be tested to ensure they work properly on the specific TDP configuration by the OEM before they are installed by the user. Finding the "blame" is proving to be difficult, as it could be that AMD is not bothering with new updates, or OEMs are not eager to test their specific configurations.   -Techpowerup. Oh by the way, that's right. That's me  you can take a look on quote below for driver version comparison between ROG Ally/ Ally X and Legion Go/ Legion Go S     Quotes Quote ASUS ROG Ally/ Ally X driver   Legion Go 1/ Legion Go S driver   What I've also noticed is that they're all on V32.0.210xx range. ASUS is on V32.0.21013 and Lenovo is on V32.0.21020. On AMD Radeon Software, the ROG Ally Z1/ Z1X driver version is 25.10.13.11, which was released on July 21st 2025. If you use Legion Go/ Legion Go S and are using that driver, let me kno
L
Ссылка
click to show
reach. So after that, I got rid of that Laptop by giving it to other member of family. Ever since that I used it for uni stuff, I even use it to draw and write essays for uni assignments. So that ROG ally has a LOT of story.   AMD Radeon prematurely ending support like this, AND NO, I am very sure they KNOW this thing is just two and half years old, them ending support like this out of nowhere, is really insulting for me. I've watched them doing this for way too long, specific on how long they support their drivers, and it's utter garbage. GCN - Specifically GCN2 and 3, especially Fiji: An otherwise still usable architecture at its 6th year? End of support. Polaris/ Vega: 2016 - 2018 architecture in 2023 while NVIDIA has a 2014 architecture (Maxwell) that at that time is eight years old, and still supported at that time? Nah, lets kill the support slowly on that year so people running RX580 and 590 on their system can just coping with watching their Spiderman 2 either crash or runs like ass than the 'competition' (no it's no longer competition at that point. It's total domination). RDNA/ RDNA2 incident: 2019 and 2020 architecture, still sold widely and used widely? Lets not give them more support and use 'maintenance mode' as a cover up. By the way Maxwell was supported for almost 11 years and Pascal was 9 years, at the time they announced it.   And now this. You see, I was expecting 10. TEN YEARS of support for this thing especially that it's on RDNA3, which is still 4 years old by the way, and this is also a gaming handheld specific APU. But no, Radeon has to pull their 'but I'm just the underdogs I can't do long driver support waaaaa' card and slaughter every single bits of confidence I have with their integrated graphics solution. And no, they should also not hiding behind the Linux community and say 'just use Linux bro it's still supported there'. Community support can only goes
Open in Telegram Каталог площадок Искать в ChatCrawler

A snapshot of an open public feed from the search index ChatCrawler — “Google for public Telegram”; refreshed as the venue is crawled. Times are UTC.

Public content only, official Telegram API. About · FAQ · What we do not do · Remove a page · Catalog