.Apex.

I just meant in the sense that you run out of VRAM beyond those optimizations, there's no way around increasing or substituting your VRAM, even though the GTX 10 Series suffers from terrible FP16 performance, but otherwise it's a good point

Hmm, it would've been 2 years ago, but considering the GPU shortages I think that sounds alright? To put it in perspective the GTX 1080 is around the performance of a regular RTX 2060 which costs around the same when new

GTX 1060 should be "okay" but as said before it depends on the usage, the most important aspect is the VRAM, a lot of midrange GPUs even from that era are fast enough for machine learning, but if the model doesn't fit in the VRAM then it's useless and there's no workaround for that, unless you train on the CPU which is much slower Don't use any of the AMD cards as their support is severely lacking for machine learning a GTX 1070 with its 8GB of VRAM is also a decent choice, if you can find one for a decent price

Oof, that's terrible, and it's easy to mistake it for being legit because of its generally appealing layout, but the file was instantly flagged by Windows Security as severe, I've tried it out on Github first and that ran through just fine, even though I didn't need it as I'm already on Windows 11 Though thankfully the Github link is what appears first on Google

Awesome work on this program, way more informative than the mess Microsoft released

Try disabling Delivery Optimization (Allow downloads from other PCs)

Those are not capacitors, they are Inductors/Chokes, seems like they used thermal paste on it, you could use thermal pads instead or any non-electrically-conductive thermal paste, though it isn't necessary to apply as inductors rarely overheat especially for a GTX 1050.

Do you have a Tesla I can borrow? I might or might not return it

Another thing to point out is that Autopilot and FSD (Full Self-Driving) are separate functionalities and subscriptions, and it was not mentioned whether Autopilot was purchased or not, Autopilot has a couple of features like Traffic Aware Cruise Control which sets the speed based on the cars in front of you, and Autosteer which keeps the car in-lane, considering how old these people are, is it possible that they've mistaken that for FSD? Edi: Wait, Elon did mention Autopilot was not enabled, and it would requires lines to be enabled, nevermind

That's fair, I also found a piece of information that suggests there was no 3rd person involved "Police told The New York Times that minutes before the crash, the wives of the men involved heard them say they wanted to go for a drive and were talking about the vehicle’s Autopilot feature. The men were 59 and 69 years old."

He only mentioned the passenger side being mangled, that's why I mentioned the possibility of the driver side

Hmm if there was no driver and the passenger on his side and the passenger seat are mangled, could it be that the driver survived and fled to not deal with repercussions? as he might have been driving erratically, or the death scene was overwhelming If it's that then they should have been able to find who the car belonged to, though I don't know much about the story other than what you posted

At stock speeds it should be fairly easy to cool This is a great option for the price https://www.amazon.com/Quiet-BK008-Pure-Rock-Slim/dp/B01KVNCEIG Can't you wait though for your other cooler to arrive?

You're welcome, I'm glad it's working!

In my system with an 8700K I cannot for the life of me manage to tame the temperatures that Prime95 Small FFT generates, and it's mostly because my CPU has thermal paste between the die and the IHS, so I stopped caring about reaching that type of standard, the fact that your CPU is under reasonable temperatures at that load is way more than good enough.

That loop temp is about average for liquid coolers, and you're right in some sense that the temperature difference between the loop and the CPU would indicate the heat transfer efficiency, but I would expect that from a Zen 2 CPU as the die is much smaller and therefor the heat is concentrated so it doesn't dissipate the heat as well, you might improve that if you used better thermal paste but I doubt it'd be affected by much, this cooling performance is inline with other liquid cooled systems I've seen using a Zen 2 CPU. Though I should mention that Prime95 Small FFTs is extremely unrealistic and you would never reach those temperatures in real world applications, I suggest you use Cinebench instead, and since Small FFTs pushes your CPU to the limits then it might be that once again because the die is smaller the CPU IHS cannot dissipate the heat as it's saturated.

Assessing idling temps are difficult without measuring the power usage in each scenario, I'd look more into the maximum temperature as that's more consistent to measure and having a loop temperature of 35C sounds about right, it shouldn't be as hot as your CPU as then it would be worthless as there's no temperature difference. I can't tell you if the AIO is working optimally or properly but it's definitely working, if not then you'd thermal throttle instantly. Comparing a 240mm AIO with a single-tower aircooler should be similar in performance, I think..

That's strange, I've never heard about incompatibility issues with a Gigabyte GPU and MSI Afterburner which makes me think It's either driver issues or a BIOS issue. Try using DDU to completely uninstall all the GPU drivers on your system and reinstall the newest ones and try downclocking again, and while you're at it, download GPU-Z to get the BIOS version of your card and export the BIOS file and upload it here, and take a screenshot of GPU-Z's main window as well. DDU: https://www.guru3d.com/files-details/display-driver-uninstaller-download.html GPU-Z: https://www.techpowerup.com/gpuz/ Saving the BIOS in GPU-Z:

Or does reverting to stock memory clock only happen at reboot?

Even if you keep MSI Afterburner open it would still refuse to reduce the memory clock? hm sounds like then there might be a custom BIOS What's the model of your GPU? you could try the brand software instead for overclocking

After a restart it should revert back to stock speed unless you've made MSI Afterburner to apply your custom settings at startup, though 6800Mhz might not have been enough, try turning it down to 6000 and reduce your core clock speed as well to possibly 1850-1900, if it still causes artifacts then the VRAM or even the GPU die are most likely defective

It would not be a significant upgrade CPU wise especially not for the amount of money you're going to pay to fully transition to an AMD CPU, I suggest you wait for the next Ryzen CPU and buy an RTX 3080 instead as an upgrade if you can get your hands on it. Between the 9800X and 5800X there's a 25% Singlecore improvement which would be significant if you were to raise the amount of cores, but extra cores in your case wouldn't affect gaming performance, and you could lower that gap even further if you were to overclock the 9800X which would put it in the range of 10-15%.

Looks like VRAM instability artifacts which usually occur when the card has been used for a long time which wouldn't be the case for a 2070 Super, or that it has been stressed or heated consistently for days such as the case for mining or it was simply just in non-ideal conditions inside a case, though it could also be that the owner received it defective like this. Though it also might indicate that the VRAM frequency is too high, a miner could have flashed a custom BIOS on it, can you check what your GPU Core and VRAM frequencies are? regardless if there is a custom BIOS or not, you could try downclocking your GPU VRAM slightly if it has degraded.

The card does have a "FanStop" feature, and it's called 3D Active Fan, and it seems like you can disable it through their Aorus Engine software I don't think the GPU has a BIOS switch, it's not listed on their webpage and manufacturers don't usually mention that which is annoying, but if you do find it, you can try switching it while the system is off

I'm talking about the BIOS switch on the GPU itself, which you haven't mentioned, but if it's an ASUS GPU then I think it does have a physical BIOS switch What's the model of the GPU?