n0stalghia

In the same way as Panama is a North American country, yes. Kinda, sorta, technically yes.

When I cold-boot my PC (for example, after turning it off overnight), for the first half an hour after booting, the USB devices will disconnect and automatically reconnect again. The devices in question are keyboard and mouse, which are both connected to a KVM switch (ATEN CS1822). The KVM switch is fairly high-end, and the second system connected to it does not experience any issues. It is 99.9999% not a software issue since I reinstalled Windows yesterday to fix it; so I'm on an empty/clean system with no old drivers/multiple drivers issues. It persisted through a fresh Windows installation. What could be the cause of the problem? KVM switch? something with the PSU or the motherboard?

I want to point out that a statement that you don’t notice errors because you spend so much time analyzing data is contradictory. You’d think that spending more time would make you recognize the errors.

So, assuming a gigabit connection and a bad router, this is possible, right? I just ran a gigabit cable directly to the socket in the wall and had the smoothest Dota match. So it seems it is the router.

Thanks, I'll try that. What about the router however? Do you know if it can drop packets?

It happens for a quarter or half a second, then everything is fine again for the next 10s. Is there a different way? It would be nigh to impossible for me to catch it perfectly... Also, what domain should I trace the route to? The game's servers? Google.com? To the router?

I've recently started having packet loss during online games. I've been trying to find out why, and one possible cause is my router. It's a "cheap" TP-Link Archer C6. It would seem that every time when I experience packet loss in-game, the router's CPU usage shows as 100%. Could the router be "busy" and skipping packets? I have a gigabit ethernet connection to the router, and my internet is a gigabit connection in a university dorm. It's possible the university internet is at fault - there were problems before - but they fixed them about half a year go, so I would think it's the router or something else. Alternatively, how can I identify where the packet loss is happening?

So, how long until LTT rebrands. Linus has been annoyed with the "Linus" Tech Tips for several years now. Now it's a new CEO. You know Linus will push for a rebrand/rename as soon as the situation allows.

Yes and no. They (Linus and Luke) talked it about on the WAN show - but ended up with a "we have no idea, because X is impossible, Y is impossible, and Z makes no sense". They talked about the topic for 4-5 minutes and left it at that because they didn't have an idea that could explain it. So I thought I'd make this thread so people can bring new ideas that Luke and Linus haven't considered.

First of all, if there is a news topic for this, or a different topic that is better suited to discuss this, please link it to me and close this topic if required. I could not find one. Otherwise, this thread can be used to discuss theories behind the "Chinese CPUs" that appear to be Intel in origin. My theory is that those chips are most likely retail Intel chips that got rebranded as Chinese chips so they can be legally sold to Russia, bypassing the embargo. Or semi-legally, anyway. Russia will say "well look, we're buying Chinese CPUs" and China will say "well look, those are our own chips".

Ah, OK, thank you - so it turns out to be marketing BS. Gotcha. KabelDirekt seemed like a reputable company, I have a couple of their shorter cables and they do indeed only sell HDMI 2.1 stuff up to 5m - everything after that is active. But good to know nonetheless, I will look for active cables only. EDIT half a year later: Sorry to necro an old thread, but I thought I'd leave a message about how the situation turned out for me. It was indeed a faulty cable; I ended up purchasing a Highwings active HDMI cable based on the Amazon reviews (I looked at the one with the smallest % of 1-star reviews). It's been running flawlessly for half a year now.

Thanks for the swift reply, I was conducting some investigation in parallel to your response and have come to the same conclusion. As I've read your comment I already started the RMA process - the bloody thing has 17% 1-star ratings on Amazon all complaining about my issue and the supposed HDMI certificate is bupkis - there isn't one. The cable needs to be at least 14.57m long, sadly it's quite a long run between my desktop PC and the TV. I have four monitors, so all output slots on my GPU are used. Therefore, every time I use the TV for gaming I need to unplug a monitor and plug the TV in. That's not a big deal, but I was worried about the cable being damaged over time if I keep folding/unfolding it on a daily/weekly basis, so the idea was to do one cable run along the walls and just leave the cable there. Less risk of damage this way. If I would fold/unfold it after every use, I'd need a ~10m cable. You wouldn't know of a reputable HDMI cable company in EU by any chance? I can only find "Zeskit" as a recurring recommendation, and I don't think their cables are sold here in Europe.

I'm trying to get my Sony XR-A90J to work with my PC. It started off quite well: I bought a HDMI 2.1 cable for 30 EUR from a no-name brand, connected my GPU (RTX 3090, latest drivers) to the TV and everything was working. 4K, 120 Hz, HDR. I'm running Windows 11, if that makes a difference. However, after about ~20 hours of use, the TV started to randomly flash to black for 2-3 seconds when playing games. I assumed that it was due to the cable, as I've been rolling/unrolling it after every gaming session and it was a no-name HDMI cable. I just thought it broke/was cheap. Both the audio and the video were gone for those 2-3 seconds (I'm using the built-in speakers on the TV, so the GPU is outputting both the audio and video signals to the TV). So I bought a more expensive cable: active optical HDMI cable from a well-known German company (KabelDirekt) that has a HDMI 2.1 certificate. It should absolutely work. And yet, as you can see in the video, it doesn't. With the new, expensive (80 EUR) cable, the screen constantly turns on and off. This causes both the audio and the video to stutter/disappear. I've tried changing the HDMI signal format between Standard format, Enhanced format, and Enhanced format (VRR). In the Standard format, the picture maxes out at 4K/60 Hz in Windows settings, but it does seem to be stable. However, HDR and higher refresh rates are not available. Has anyone had similar issues before? I've seen several posts on Reddit and on Nvidia's forums claiming it's an issue with cables, with 3000 series cards, with drivers, with VRR, with the TV chip, with the TV firmware, but that's like every single variable in this equation... Relevant data: RTX 3090, Windows 11, latest Nvidia drivers v531.79. Sony XR-A90J, latest firmware. KabelDirekt HDMI 2.1 cable, 20m, active fiber - allegedly certified by HDMI.org.

I think you may have misunderstood the translation. The entire point of the article is him asking someone to step up and verify it. He's a researcher, not a startup generator - his goal is to create a new tech and publish a paper on it. They cannot do that if they can't afford a server farm that can parse the internet like OpenAI can. Is that clear enough now? Hence the second quote - they cannot run/train it on the same scale as OpenAI. Given localized scale - i.e., let's say 1000 texts each 10.000 words long. You feed this data to ChatGPT-4 and their model and ask to, let's say, write a summary. On this localized scale, their model beats ChatGPT-4. But they want to verify that on bigger datasets. Hence the article. I hope this clears up things for you.

Feeling guilty to necro an old thread, but it has to be done due to xkcd 979. Not sure what the problem was, but removing the CMOS battery and re-inserting it again fixed the issue. Then I updated the BIOS to a newer version just in case, and the problem did not reappear. So all is good. RAM speed, Secure Boot, etc., were not a factor. After CMOS clear I tried booting with several variations of those settings, the problem did not reappear.