Craftyawesome

Modern CPUs/GPUs will boost depending on temperature, so it is quite possible they are using more power and creating more heat energy then they were before. This is generally not considered throttling until the temperature and clock drop becomes excessive. I don't think "20 C worth of heat" is the right way to look at it. Larger objects take more energy to raise their temperature by a fixed amount. Think of it this way, if a CPU is drawing 100W and it is maintaining the same temperature, then 100W of heat is being dissipated. A better cooler can do this with a smaller difference between ambient and CPU temp, but it is still dissipating 100W.

There are some Ethash coins left AFAIK. And some of that hashrate is ASICs and not GPUs.

I tried the memory and disk tests in both supportassist and the pre-boot diagnostics and they passed. Memtest86 passed. I'll run memtest86 overnight. If I can't get an error, what do you think is the most likely cause? Mobo, RAM, or the drive itself? Is there any chance the slot is bad? I can try swapping the drives and see if things go worse with the primary drive in the second slot. Also, how often did it happen on the ASUS?

I haven't found anything obviously related other than "The system has rebooted without cleanly shutting down first" which is obviously not the cause.

Hmm. I hope I don't have to RMA the whole laptop. And to clarify, the drive is already secondary, so I don't think a live usb would change the symptoms.

I couldn't move it with the screw down, so that's probably fine.

I couldn't see any pins, but I removed/reinserted it anyway to see if that helps this time.

A linked it. But if you want a screenshot

I currently only have 55GB free, but this has been happening while I had >100GB free. I think it is counting leaving some sort of idle state as a cycle? The included drive is supposedly over 51k. I did remove/reinsert the drive earlier and IIRC it felt snug. Although that did make me think of something. The laptop doesn't include the mount for the second m.2 slot. And instead of going through dell support to probably have to pay anyway, I just got a third party one off of amazon. Could the tolerances be just barely off? Maybe I can try to remove the drive and try to shift the mount a fraction of a millimeter closer to the slot? I did update the firmware earlier and run all the diagnostic tests in magician (all passed). It has 8, but only two are non-zero. (It matches what HWinfo shows but in kelvin) smart.csv

My Samsung 980 has been dropping out on my Dell G15 5515 laptop running windows 10 64 bit. What happens is that my drive randomly stops responding. Explorer locks up if I try to use it and so will other programs expecting to use the drive. It also prevents me from shutting down with the windows menu because it will either get stuck on the shutting down screen or it won't even show that screen at all. I've gotten it to happen somewhat consistently a couple different ways. Sometimes steam game updates that have a lot of disk access will cause it. The other way is that it happens after a few hours of leela training (mostly random read workload). It has also happened with no/little foreground disk usage, but maybe steam was downloading in the background. On the last post I made, I eventually found that hwinfo showed a second thermal sensor (likely the controller temp) that I saw as high as 99C, so I assumed it was just thermal shutting down. So I got a heatsink, but now I've seen it stop working as low as 90C. So I don't think it's a thermal problem. Is there anything software related I should try, or is it just a bad drive? I've tried: Physically unplugging and reinserting the drive Trying the chipset drivers from both AMD and Dell All 4 diagnostic scans in samsung magician. All pass. Updated drive firmware Updated BIOS (1.6.0) Reset bios to optimized defaults Running the longer scans while downloading a game update. The percentage gets stuck, and the eta decreases until it gets stuck at 0.

For anyone who finds this, I suspect I found the problem. HWinfo reads 2 different sensors, and the second (probably the controller) is getting really hot under load. I guess it doesn't shutdown in a way that communicates that windows/programs can handle. Seems I probably need a ssd heatsink. Edit: Still happening with lower temperatures. Will be making a new post.

I think that must include leaving some sort of power saving mode. All of my drives seem higher than what is realistic otherwise. The boot drive is actually higher and works fine.

My Samsung 980 has been dropping out on my Dell G15 5515 laptop. What happens is that my drive randomly stops responding. Explorer locks up if I try to use it and so will other programs expecting to use the drive. It also prevents me from shutting down with the windows menu because it will either get stuck on the shutting down screen or it won't even show that screen at all. The only way I have gotten it to happen somewhat consistently is steam game updates that have a lot of disk access. It has also happened with no/little foreground disk usage, but maybe steam was downloading in the background. It has also happened without steam running at all. Is there anything software related I should try, or is it just a bad drive? I've tried: Physically unplugging and reinserting the drive Trying the chipset drivers from both AMD and Dell All 4 diagonistic scans in samsung magician. All pass Updated drive firmware Updated BIOS Reset bios to optimized defaults Running the longer scans while downloading a game update. The percentage gets stuck, and the eta decreases until it gets stuck at 0. Watching cystaldiskinfo and pressing F5 while downloading an update. I have seen it stop working lower than 55 C.

FWIW, only the OLED with the new dock can do 4k60 at 4:4:4. The older switches/docks only have 2X displayport 1.2 lanes which is converted to hdmi 1.4.

To add a little more info, AFAIK the versions listed in the OP fix the LDAP attack vector. Unfortunately, this isn't enough.

Maybe try enabling dynamic super resolution in your nvidia control panel to try running the game at 1440p first.

Doesn't twitch have a low max bitrate? Streaming most games at 1440p60 isn't viable on any encoder. Is even a 770's encoder still better than a modern AMD? There were significant improvements for 1600/2000, and I think improvements before.

HVCI performance is AFAIK the best explanation given so far, but it doesn't explain the inclusion of Zen +, since GMET (AMD equivelent of MBEC) wasn't added until Zen 2.

6th and 7th gen also have built in TPM 2.0.

Summary Previously, installing or updating Windows 11 in a VM would ignore some requirements like TPM 2.0. However, this is no longer happening with beta build 22000.194 and dev build 22458. Attempting to update will throw an error. Quotes My thoughts It's a shame that MS did this, since a VM is a convenient way to try Windows 11 preview builds. The TPM requirement seems particularly annoying on a Windows host. Hyper-V works, but needs Windows Pro. I couldn't find an indication that Virtualbox supports TPM 2.0 (the above mention of Oracle seems to be about Oracle VM Server or something). VMWare does support vTPM, but it wants an encrypted VM, which the free version cannot create. Outside of Windows, Parallels and QEMU support TPM 2.0. Sources https://www.xda-developers.com/windows-11-vms-will-likely-soon-require-tpm-2-0/ https://blogs.windows.com/windows-insider/2021/09/15/announcing-windows-11-insider-preview-build-22458/ https://blogs.windows.com/windows-insider/2021/09/16/announcing-windows-11-insider-preview-build-22000-194/ https://techbuzz.asia/2021/09/16/the-latest-build-of-windows-11-is-offering-an-error-message-on-some-computers-that-do-not-have-a-tpm-2-0-chip/

It looks like that because the guy who made the iNES header worked at Nintendo. https://www.resetera.com/threads/tomohiro-kawase-mightve-been-hired-by-nintendo-to-put-rom-headers-into-vc-updated-dec-1-2018.64755/#post-13593223 and at the very least, Clu Clu land was sourced from their own archive. https://www.resetera.com/threads/tomohiro-kawase-mightve-been-hired-by-nintendo-to-put-rom-headers-into-vc-updated-dec-1-2018.64755/page-3#post-13620946

It may also be worth noting that most of the games on N64 VC are popular. These games usually work well on PC emulators. Unless the less popular games on VC have way more bugs on PC than I'm aware, even GlideN64 on release was significantly better. I wouldn't be surprised if some games were better on PC in 2006 (noise emulation in sm64 teleports/vanish cap, star fox being less buggy AFAIK). They intentionally exploit bugs in the Wii VC emulator to beat the game faster. It makes way more sense to fix dolphin than to intentionally break PJ64 to match Wii VC. Also, "preferred method to run" is a stretch considering runs within 10% (OoT) or 15% (MM) must be on original hardware. Also AFAIK gz or kz (practice roms) are typically used for practice. Dolphin would probably be best for debugging a route idea.

This isn't true. Dolphin is good at emulating how it played on wii, but that experience isn't great. There are tons of minor problems, and some are major enough to be added to the dolphin wiki specifying that they are not dolphin bugs. PC N64 emulation has also improved. GlidenN64 came out in 2015 was significantly better than any HLE plugin before it. Angrylion RDP Plus came out in 2017 and is very accurate, but needs a strong CPU and is at original resolution only. And last year, ParaLLEl was released and allows angrylion accuracy and higher resolutions with a somewhat modest GPU.

They definitely aren't just using videos, demos are way too long and high resolution for that. A good chunk of demo readme files will suggest a then-recent GPU. You can also just check your GPU utilization to see it being used. As for other things that can be precomputed, they would probably just use whatever is smaller. I would assume that the majority of the time it would be smaller to use code to generate things, though. I think you are probably underselling the difficulty. Assets are big compared to demos. You would want to write code to generate what you want after launching.

I mean, they are clearly willing to support some but not all 7th gen. Not to mention zen+ doesn't support it but is on the support list.