tarfeef101

"They wouldn't" - idk man the fact that they (read: Linus) has a tendency to knee jerk react to things that merit time and care is kinda the point They definitely shouldn't say anything more for now, though. Get all the ducks in a row, and reply when you have irrefutable evidence and a prepared statement which lets the community know either "you were in the right" or "you were in the wrong, you're sorry, and are fixing it". Then make more tech videos.

Made previous responses here and here. I also replied on the FP post (but the formatting tools there still suck). I hope the feedback of a vessel-era supporter who has spent quite a bit of money on LMG, and watched every video for probably a decade+ merits someone on staff at least reading. That's all I ask. Things the above don't address: - it sounds like you ARE taking feedback after being smashed in the head with a baseball bat about it, and slowing down. This is my #1 ask, to take the time to make quality, correct content so I can trust information again and not just watch for entertainment. I hope that the time taken is put to good use. Full details linked above. - Your "new info' about Billet thankfully doesn't change my comments at all - The tweet(s) by Madison I just saw are pretty bad. I more or less believe innocent til proven guilty, but also don't require a court of law to arrive at my own evaluations of guilt. Most stuff like billet don't matter much, and need a pattern for me to take that and say "you suck, game over, goodbye". If the Madison stuff holds up (I've just once read a tweet at 5am), that's a bigger strike. Take the time, be careful, apologize as appropriate, and don't get into ugly public blame games. I watch you for tech videos. That's what I want to see. When it comes to "drama" all I want to see is "as little as possible" and "when you have an actionable item, you do the right thing and don't repeat mistakes". That's it. People have focused a lot on billet labs and similar small incidents. I care less about those things. I care the most about quality content I don't have to audit as I watch and can "just trust". That's not been a thing in a long time. Quality hasn't kept up with scope, and that's the biggest issue I have. It isn't enough to make me leave, as I haven't yet, and I didn't need a GN video to figure out you have these issues. But it was enough to make me unsub from Jay a long time ago. So I do hope it doesn't get to that point. Sounds like you're gonna make serious efforts to prevent that though, and I look forward to it. - 6 nines is a crazy goal Luke, good luck (I'm an SRE, I know how hard that is. Even just "having a cloud provider" can auto nuke that SLO, a recent lambda outage did a number on me where we'd have failed that measure for sure) - markbench open source (and testing methodology) will be great. I have professionally worked on these things when I was at Intel. I care enough to give feedback, and hope I get the chance to do so. I want my local tech media company to succeed.

lol I don't think anyone is trying to ruin a company, GN doesn't have that kind of reach, and the LTT audience doesn't care enough about the issues raised for that to happen. I DO think it's a good thing to call out their technical issues now that labs is being marketed in almost every video as a solution to those problems that being said, I am still satisfied with my status quo of "ignore or at least don't take technical data at face value", and just enjoying the content as entertainment

I see. I'm lazy and love jenk, so if I had to not use the original panel, I would: - put compressible rubber strips in the top+bottom of the mounting area (looks like the panel is inset at the top+bottom relative to the frame, whereas on the sides it sits on top of the mounting area) - cut the panel such that it fits perfectly horizontally, and is slightly taller than the new vertical height w/ the rubber strips glued into place - friction-mount it in - you can probably just use thin double-sided tape on the sides if you want (would also look nicer that way if you used coloured tape to hide the metal tabs underneath) for more security side-to-side

The easiest way is to re-use the existing side panel, and cut out a large rectangle of it to replace with acrylic.

https://www.techpowerup.com/gpu-specs/asrock-rx-6900-xt-phantom-gaming-d-oc.b8530 listed TDP is 300W. Calculate how much power the rest of your system should draw, and see if it is less. Keep in mind that sometimes load transients (short changes in power draw and voltage as load changes) can create temporary periods of load exceeding the rated TDP/power draw of a component, so some buffer should be used. This GPU (unlike, say, the nvidia 30-series) isn't famous for huge transients, so with a CPU that draws under 200W, I suspect you'll be fine.

I mean I'm capable of independent thought, so I have noticed these quality issues for years. I am still around because I get enough value watching LTT for entertainment instead of technical information. That continues to be the case, though I do hope they improve those quality issues. The "main thread" has further insights for those who care.

IF it has a boot partition (often shown as EFI in windows), it might cause your motherboard to boot to it instead. you will want go to your BIOS and set the boot order to ensure you boot from your main drive once you're in windows, one of two things will happen (well maybe 3): - the drive is empty, and you'll be prompted to format it. then it'll be like a new empty drive with a letter - it will be pre-partitioned with filesystems. you will have access to your old data, and can choose to keep it as-is, or reformat/wipe it with disk management - it will be corrupted data, and you will NEED to reformat/repartition it with disk management

I mean yes, but also no. this is old news. but also, I don't have to care. I don't have a parasocial relationship with LMG. I feel as bad for them as anyone else w/ a boss that ignores them, but that's where it ends. In the end, I care about videos. At the very least, even if I do have some parasocial vibes, I'm trying to remove them from the equation here. Yes. And this is highly coupled to my point. Like #1, I don't really feel the need to care about their internal processes. I just want the end result. As consumers of the content, we should be focusing on the content, not focusing on giving free advice on "normal internal communication for a company"

I already posted an "overall thoughts" comment (still hoping to see an LMG set of eyes on it, I do think that someone should be reading every comment here, even if they don't reply to all of em). So I won't restate those opinions/observations. What I will say is I'm surprised (well really just disappointed) with the direction of the comments here. Everyone is focusing on the smaller issue of Billet Labs. Linus probably wasn't involved in that at all. It was a process issue (a real one that hurt Billet, I detail this further in my initial response). But it isn't "the main point". The main point, for me, is the issue I noticed (and so did Linus, to some extent) long before the GN video: quality is lacking. A few WANs ago he talked about some kind of "bug bounty for viewers" because he knew that they kept messing up. It focused more on stuff like "misremembering a part of a display specification" or smth like that, rather than "not noticing a 300% performance deviation in a test", but there was some awareness. Yet, these issues have been going on for years, at a frequency (even accounting for the rate of publication) exceeding peers (at least the ones I watch, I have dropped subs before due to frustration with misinformation, notably jayztwocents). Their solution has been "we don't need one", and production has continued at a breakneck pace, and errors are missed or fixed in post. Again, my original comment details this further, but my point to be made here is "this is the main problem, priority on speed/volume over quality, even when shifting focus to a labs-centric, data-forward branding of content". That is what the focus should be. The interview videos said it all right to the leadership's faces. They are moving too fast. But that hasn't been talked about. Linus has responded to easy, irrelevant stuff and shifted the discussion to pointless internet bickering, but hasn't addressed my concern as a longtime viewer: that I can't watch LTT videos with the intent to get technical information. I've long since classified LTT/LMG as "just for entertainment and product discovery". That is what I want Linus (and everyone else) to be talking about.

I hope this gets read and isn't just lost in a huge thread. I'm a looong time viewer and supporter. I was a vessel subscriber all the way back then (and am a proud OG tier on FP to this day). I definitely am broadly "a fan and supporter" of LTT. I have also bought GN merch, and have watched them for a long time as well, and would say the same about myself there. But I do believe I'm pretty able to separate myself from either side and be objective, so here goes: - I think they had a lot of good points. This is a conclusion I reached before they made those points, because I noticed those things myself. I frankly don't want LTT for educational/informative content. I mainly watch for entertainment, and for consumer electronics like phones/laptops on SC, insights into high level usage stuff. You guys do make a lot of mistakes. That's just the truth. - Those mistakes are often things I believe can and should be fixed by a reshoot, voice-over, etc. Something that cannot be missed. It definitely seems to me like presenters/hosts are not intimately familiar with scripts before shooting, and have limited takes, as most info errors don't seem to be wrong consistently, but in specific takes, or were mentioned once incorrectly but the conclusion seems derived from correct info. 1000% that screams "you're shipping too fast to take care". I don't mind that for Alex's latest jenk cooling project. I do for a product review. And I agree that many people like myself primarily listen, don't watch intensively. So I think you should do better there. - You do also just make errors in testing, research, etc (not just slipping up words when presenting). The skates on that mouse, I think were inexcusable. If my mouse was surprisingly rough/slow, the first thing I would do is explicitly check for plastic. How that wasn't considered is beyond me. The testing errors with game settings are pretty bad too. Data that erroneous/dissimilar should merit follow-up and be caught. I speak from experience, I did GPU validation for a living at Intel for some time. I also just generally see a lot of random errors about things like misstatements about a given specification, etc. You brought this up on a recent wan. You're aware, but haven't yet fixed it. This is frustrating for me. Getting basics wrong and telling consumers those things is what turned me off of jayztwocents in a large manner. He did that a lot and I got tired of watching someone shove out incorrect info to those who wouldn't know better. - when it comes to billet labs, I'm mostly on your side. I interpreted that video as "hey let's play with bespoke fancy cooling". I agree your conclusion wouldn't have changed. I DO think you should have just "done it right", and I recall wishing for as much. But I don't think you will have burned them down because you didn't. It reflects worse on you than them, in my opinion. I do think not sending the item back is bad though. Even if you offered money (weeks/months later), that doesn't mean they don't lose valuable time as a startup with limited funding burning overhead costs. You should do better in that regard. I'm not gonna stop watching LTT, or GN. I hope you do listen to your pace feedback from your team, because I do believe that is the cause of basically all these issues. I'd love for a day when I see an LTT review and I can just take it at face value. Right now, I don't. Some testing where methodology is public and I can tell it is done right, I'll use as info. But LTT has for a long time been primarily entertainment and about "discovering tech to research myself". If you want labs to change that to "the place I go for reliable reviews and data", then I agree with Steve overall, you do need to do better. I only skimmed ur post here (maybe 74% attention), but I recall seeing references to you justifying errors as growing pains and "it's okay because you're transparent". I hope you change that attitude. It doesn't have to be like that. Y job is helping software companies be reliable and scale. Many will use same excuse (growing pains). But I've proven you can grow reliably. It just takes more time/care, which you hadn't prioritized. Again, if you're talking "meme cooling vlog" content, those errors don't matter much. But if you're positioning as informational and data driven, they do. I think you should take that time/cost to do it right. Telling me you know your priorities are different (transparency) doesn't make me agreed with them. That's the fundamental issue here.

As someone who worked on similar systems and also had to use them at a hardware manufacturer, here's some tips/features that'll make everyone's lives better: A proper cli, and NOT an interactive one. You should be able to kick it off just with flags/args and have it run, then exit when finished Ofc the automation for games is gonna be jenk, terrible, and platform specific. Just how it goes. But the tooling around it can be made cross platform, and individual tests should be classes which can be defined for different apps and OSes/arches Rely as much as possible on OS and vendor tools/APIs vs random python/go packages to get metrics. They die, get unmaintained, etc all the time. Unless you plan to pick them up as well Don't try and make metrics in your automation tool. Just collect raw data and let the data people use it in the rawest form Doing things the "intense" way where you have a controller system running something that hooks into a server on the DUT is a lot of work, but rewards you with power cycling abilities, handling crashes, hooking into computer vis tools you're developing to analyze the screen, etc. Others do this, for a reason. Idk how big your team is or what your timelines are like, but long-term this enables a lot more data collection than otherwise available running locally only Devs probably don't get a say, but time spent making a UI is time wasted. A good cli is all you need for automation. It is better, in fact. But since your stuff is also used by a media company and maybe the public idk how much you can afford to ignore that. All I know is I have wasted many days trying to automate proprietary tools who have fancy UIs but not 1000s of people to interact with them every day to run all the tests they're used for, and I always wished their time was better spent

that assumes the internet never gets faster, and that they have the access point also on the switch, which many don't cause they use an all-in-one router, or have wireless using another subnet and port from their router. if the AP is going through the router, then this won't apply. it's future-proof, and not an incredible cost IMO

No worries. Just one thing to remember with a normal gigabit switch: not only do all the devices downsteam of the switch only get up to 1Gb, but since there's only a single 1Gb connection to the router, they all SHARE 1Gb. The benefit of something like what I suggested is that you can have 1 or 2 devices that can saturate up to 10Gb by themselves, but since there's a 10Gb uplink to the router as well, you can have MANY devices going at 1Gb or greater. Which is handy if you have high traffic, especially simultaneously.

For a consumer like you, just get an unmanaged switch. setting it up will be very easy, you literally do nothing! If you care about saving money, I'd recommend something like this: NETGEAR 10-Port Gigabit/10G Ethernet Unmanaged Switch (GS110MX) I have one myself. Easy to set up since it's unmanaged, has 10Gb backhaul and a couple dedicated 10Gb ports so the key devices that need it can be super fast, and the rest still get to share the 10Gb backhaul, so it's a great value. And also I don't even know if it has a fan. If you need more ports, and don't have multiple from your router already, just grab another cheap switch for like 20 bucks, as I suspect most devices don't need more than gigabit speeds in your home, so them sharing that should be fine

okay i noticed you're doing direct nvidia driver installation AND geforce experience, i'd recommend picking just 1. and if you pick the latter, doing so has never been a source of issues for me. Well I assume those usage numbers are under a gaming load, though idk what game (and that does matter). But "consistent" clocks does not necessarily mean "expected boost clocks under gaming load" and "usage seems normal" idk what you think is normal, but something should be more or less maxing out, you're not hitting game engine limits at 10fps, that's for sure.

Well, I mean, looking at that case of yours, 3 fans can only do so much with all that glass in front of them. Not exactly a performance-oriented case. That won't help with thermals, which with a 12900k (which on most boards, at stock has an unlimited turbo duration), you might need help. Also check your voltage, a lot of boards super-overvolt stuff out of the box. JayzTwoCents just uploaded a video demonstrating this with your CPU:

The 12900K is, simply put, a hot chip. Depending on your board settings, if you did a manual OC, the case airflow, and the conditions under which you're experiencing throttling, this might just "expected behaviour". More details are needed to know if this is actually user error or just a 12900k being a 12900k

Lot to unpack here, but I'll say this much: A general rule that's safe to follow is "don't download software from motherboard vendors" If you want chipset drivers, grab em from AMD directly. Grab your GPU drivers from nvidia. otherwise, just let windows update take care of things for you unless you're having issues after letting it run. give that a go, and see what's up. i also recommend logging device usage (cpu/gpu usage, temps, memory usage, clock speeds, etc) whilst gaming so you can identify "what is doing what it is supposed to"

Despite the above reply (which isn't entirely incorrect), some PSUs, especially older ones did used to come with an external lead for the fan within the PSU. However, anything in the last 10 years is very unlikely to have that. If your PSU doesn't have one, all that means is the PSU manages its fan independently, so you have nothing to worry about. You could repurpose the header on your board to power another fan if you wanted, that's fine. You didn't do anything wrong there. As for why you're getting beeping, you probably have a BIOS setting to warn if a fan header is reporting 0rpm (the board things the fan is broken). You should be able to adjust that

"however fast you can make it without it thermal throttling or becoming unstable" so basically, just keep testing until it becomes unstable. then go back to your last stable result.

There's gonna be opinions all over this, so forgive my brevity, but I really dislike getting into posts that turn into opinion/fanboy wars. I find that for a tech-savvy person, the best solution is a cheap, old PC running PFSense (sounds daunting but really is not any harder than off-the-shelf routers to configure), and a decent access point/points, and unmanaged switches if needed. It's simple, effective, and reliable (with the APs being your likeliest point of failure). When you set something like that up properly, the idea of "oh the network is acting up, restart the router" simply doesn't exist, and that's the best feeling ever. For that PC, I'd recommend: - at least 2 ethernet ports, 1Gb or higher (one for WAN, at least one for LAN) - 2GB of RAM - 2 cores from 2012 or later? - really doesn't need much power, tbh. For the switch, just as many ports as you need for ethernet devices. ones like what I have which have 10Gb backhaul but many 1Gb ports for downlink are great because you can have many people saturate 1Gb at the same time. But probably not needed in most homes For the APs, this is the big "opinion" section. Lots of fanboys of every company. I personally like TP-Link's Omada series of Access Points, they're cheap, not hard to configure, support all the fancy features I'd want like seamless handoff, and have been reliable for me. I'm sure you'll get Ubiquiti ppl in here soon, and probably a couple others. Do your googling, pick what you think works best for you. I'd just say "avoid mesh networks, and avoid consumer-grade all-in-one devices". The former are inherently less good than multiple APs which are managed by a controller and have their own ethernet connections, and the latter just die and/or aren't reliable, most of the time.

GL lol i'm the opposite, i'm too cheap to buy the fittings necessary to create easy drains, and tend to avoid builds that are "easy" or "normal" so I always seem to have terrible draining options. but i do leave myself fairly accessible fill ports. except for my testbench, i'm very proud of the fact that I actually built that properly for once

If you can't tilt it (which is pretty not great), the best (though not ideal) option I can think of is to attach a tube somewhere above the water line in your res (or elsewhere in the loop) that's still just air, and blow air through it to force the water through the loop. do this until there is enough water going through that the pump isn't running dry, and you can top up wherever needed. And maybe in the future design your builds with the ability to fill them in mind

Um... why are you trying to use an adapter like that? The RM1000i has enough cables included to not need adapters. I think we need a little more context on what's going on here. Btw a general rule of thumb is "avoid splitters, and use as many 1-1 cables (not daisy-chained ones) as you can"