Addvilz

And that's totally fair. Creating hardware from scratch costs pretty penny. Might have been a bit more transparent about it.

Why do these look awful lot like Startech/whoever actualy makes them KVM switches? For example, these two: https://www.store.level1techs.com/products/p/14-display-port-kvm-single-4computer-6zepx-s4kf3 https://www.startech.com/en-es/server-management/sv431hdu3a2 Like, they are not similar, they are identical (case, not specs ofc. because I am too lazy to search for one with the same specs)

Disabling Precision Boost seems to have resolved the issue, no issues so far. FYI for whoever might also might be experiencing a similar issue. Thanks everyone for help.

Been running stress on CPU an memory for the past 30 minutes. No issues of any kind will try to run the current config with PB off for a while and see if anything pops.

Yep. So nothing more I can add here.

You could try to do disk write speed test with https://crystalmark.info/en/software/crystaldiskmark/ Just to rule out a damaged drive. But that's about all I can offer here.

Same thing as I already mentioned, your speed test results and your download results can be wildly different. And it does depend on many factors like where you are downloading from and who you are download from, even time of day - how many other people are downloading stuff. Even COVID could be affecting stuff now since a lot of people are downloading a lot of stuff - all the time. Given the metrics you gave me I don't think there is an actual issue there. Sure it could have been faster for some downloads at some point but who can tell. Your disk use is very low and 970 EVO is a decent drive. Your speed test results are as expected too. Unless you can give me a bit more details about how the "very weird" is manifesting there isn't anything that I can recommend you investigate further.

Are you perhaps mixing MBps and Mbps? They are two different things. Your download speed in the screenshot is 34.2 Mbps, or around 4.2 MBps - same as write speed on that SSD. Pulling 34.2 Mbps from Steam is okay-ish. I can pull ~50-60 Mbps at times, and I have a symetric gigabit fiber. VALVE just does not seem to want to allow you to eat all their bandwidth.

@the1collin This actually sounds more like hitting a hard drive / cache / SSD write speed cap / limit / something. Check your disk use % in Task Manager (assuming you are running Windows) while downloading. You might also have some issue with AV software. Just try to pull something large from steam and see what happens in Task Manager. Web speed tests pretty much never hit persistent storage, only RAM and CPU. So you could see awesome speed for streaming and on speed tests, but actual downloads might be capped by the speed and caches of your persistent storage - HDD, SSD etc. @The_russian on a fast lane +/- 10Mbps is within reasonable margin of error. Relatively few ISPs seem to do this, but nevertheless. I don't trust Ookla any longer. I have seen contracted 600/600 symetric fiber showing 800 on Speedtest, and ~350 on Fast / UI. I did in fact used to host my own "speed test" software loooong time ago just to get consistent measurement because of this reason. I don't do networking nowadays so no need, but.

A little bit off-topic, but ISPs are known to cheat with Ookla speedtest and route traffic to it over significantly faster lanes you would get for all other traffic. I'd avoid it. Fast.com is a bit trickier - it runs on Netflix infra and ISPs would need to whitelist all Netflix traffic to influence those speeds. UI speedtest is generally limited to sysadmins and ISPs seem to be largely ignoring it. I have personally seen Ookla speedtests exceed reality by anywhere from 20 to 50%.

Your network speed seems to be oddly asymmetric, but the speed is just fine. The problem is most likely with the game servers themselves or the physical distance between you and those servers. Your network is unlikely to be at fault here since two speed test tools I have personally used countless number of times and trust have given you very respectable results.

That is a fairly terrible way to test network speed. Try fast.com or speed.ui.com. This should give you somewhat more reproducible results. Try the above from multiple wired and wireless devices. If they all show similar results, your problem is methodology. Nothing to do. If you see significant difference, first try swapping ethernet cable between your PC and router/switch/wherever it is plugged in. Try switching between different ports on the router (if possible).

@Fasauceome I am not entirely sure I want to screw with downgrading BIOS just yet. Perhaps with all other options exhausted. @svmlegacy disabled PB. I guess we wait and see.

@svmlegacy memfreq is at 2400, this is the exact RAM stick installed - https://www.crucial.com/memory/ddr4/ct16g4dfd824a. I'll try to disable Precission Boost and see if that helps. Might be relevant, might be not, but the system runs Ubuntu 18.04 (well Elementary OS technically), latest updates. I did some stability testing on it (CPU and GPU load) a while back and nothing immediately crashed either.

Have a problem with GA-AB350-Gaming 3, BIOS version F50a. Without obvious reason the whole system would lock up and become unresponsive Display output becomes blank (no signal) USB ports sans power output CPU and DRAM status LED's are flashing sequence CPU, DRAM, CPU, DRAM, ... Hard power cycle is the only way out Happens randomly, sometimes once or twice a day, sometimes a week in between This is relatively new, observed for a month or so Has happened with the system on idle and under heavier load No significant changes in hardware immediately before the issues started No overclocking whatsoever, changed fan curves a bit to adjust for noise and enabled virtualisation, the rest is default BIOS settings In so far I have tried without any luck: Reset the BIOS Re-seat the CPU (it was time to do some cleaning anyway), no bent pins or visually identifiable issues Swap the RAM between slots Run a different GPU https://download.gigabyte.com/FileList/Manual/mb_manual_ga-ab350-gaming3_e.pdf According to the manual certain indicator LEDs indicate issues with particular hardware but I do not see any mention on sequence of LEDs flashing intermitently. No clue what the issue could be. System: Ryzen5 1600, single stick of 16GB Crucial RAM, Corsair VS550 PSU, Corsair MP510 NVME, XFX RX570, +3 drives on SATA, Intel AX200 PCIE WiFi card. Clues?

I am not worried about the stand as much as I am worried about the glass surface.

Would building a custom enclosure be another potentially reasonable approach that would not require to maim the hardware? Some wood, noise insulating foam inside panels, zip ties kind of a thing... Provided sufficient air intake and exhaust routed through the "rack", I suspect it might work.

I have 2x EMC 1U servers with S2600GL Intel board, each 2x Xeon E5-2640 (LGA2011). Each server has 6 chassis fans + 1 PSU fan. The chassis fans are connected using what seems like a proprietary 10-pin header on the mobo. The issue is that noise from these machines is completely unbearable. Think, apartment building, thin walls, the works. Hiding the machines somewhere is not an option - no really appropriate spaces for that. The only way out I see is tone down the noise. What would you suggest be my options realistically? Is there any CPU cooling solutions that could handle the heat AND fit in the 1U chassis, and bring the boxes down to near-desktop like noise levels? Is water-cooling the only viable option here?

No mention of ElementaryOS? I am disappointed, @RILEYISMYNAME https://elementary.io

No issue whatsoever doing that, as long as you mentally keep those languages boxed and don't force one to work like the other one. But I don't see why not.

So the autocomplete does not work for any project or just that specific one? 1. Try "restart and clear caches" menu item. 2. Are you happen to have entered power save mode in IntelliJ somehow? Try see if that's enabled. Power save basically disables all heavy functions. Including autocomplete.

All aboveplus Sanctuary Expanse Killjoys Dark Matter Farscape Warehouse 13 Defiance Orville Firefly Terra Nova Fringe ... to name a few ?

Can you elaborate?

+ glances (find it in pypi) and lm-sensors for sensor data. Not sure what you are trying to do though.

It is not possible. No, yes, I know, but still no - it is not possible. And the reason for that is, there is no "original text behind there". Images do not work this way. Whatever was there was replaced with the black blocks and that's the end of it.