Paul Vreeland

Black Desert Online is a CPU bound game. Notice all that physics, especially in highly populated areas? (Player clothing and hair, boob jiggle, Horse clothing and hair, Wagon accessories) In BDO, all physics are currently handled by the CPU. If you are in a populated area, your FPS will tank due to all the physics calculations. They are not offloading that work to PhysX or other Compute language in the GPU, all in CPU. When your CPU is too busy to handle the load, delivery of data to the GPU is delayed. (100% CPU / lower GPU %) You cannot turn down or disable physics in BDO. You can use low poly avatar models I forget the setting name where Faraway people are like black shadows, NOT "high end mode" (keeps player and npc render distance lower). But get more than 20 people and their horses on the screen and FPS WILL take a hit. It doesn't help that most players play females with the added boob jiggle and more jiggly outfits such as skirts.

Teardown and rebuild. You could be shorting something somewhere. Also, you may have damaged a component during case transfer. Diagnostics also help sometimes.

No moving parts last forever. You can try giving them a different power source if the bearings are't worn out.

You can't make assumptions like that when troubleshooting. If you are legit hearing electrical arcing noises, that thing needs to be looked at by a professional. Or you can tinker around with it until it blows something and requires complete replacement of nearly everything... Popping noises on the other hand can be a myriad of things. I have mostly noticed pops coming from sound devices picking up interference such as a voltage surge when powering on or electromagnetic noise/interference coming off other PC parts.

BACKUP BACKUP BACKUP. If you have anything you value on that PC, keep it in a second location. Second location, not a separate location. Two places minimum. Suddenly noisy HDDs can be an issue completely unrelated to what you are having and can result in data loss. Depending on the configuration of the system, that HDD could be a factor in your symptoms. Starting with basic tests and visual inspection can reveal other issues as well. If I've learned anything in the last 16 years, it's that you don't want to just check and fix 1 thing. If you are already in the machine, give it blanket testing and check as many things as you can. If you aren't a technician, you may not know what to check unless you have some experience. Without "peeking under the hood", it's hard to say what the problem is. There's "a million things" that all need to work perfectly for you to push power button and enjoy the computing experience. One thing fails in the chain, one tiny component hidden away on a circuit board, and the chain fails. You can always call a technician, too. There's a reason people pay us money to get stuff fixed, it's usually after they waste their weekend, month, year trying to hack at it themselves. I'm not saying don't try, but this is a realm where the problem isn't easily reproducible so it sometimes requires a keen technical eye to spot the issues.

If you aren't losing settings, it is not the battery. That doesn't necessarily mean the battery doesn't need replacing. 3 years old isn't fresh. As a technician, this is the kind of case where a teardown/visual inspection is a good starting point. I usually check for swollen caps, corrosion, burn marks, liquid spills on the motherboard. Then move into testing the CPU/RAM/HDD. Yes, I said liquid spills on a desktop PC motherboard. Seen it in the past...usually soda. Updating UEFI/BIOS might help, but something working one minute then not working the next generally leans towards component failure. This kind of situation is usually diagnosis of exclusion. Sometimes the PSU can be at fault, sometimes motherboard, cabling, dust, etc.

Always test a new build before putting it into production use. CPU/HDD/RAM...make sure they pass tests. That would be something like Prime95/Gsmartcontrol/Memtest86+ New out of the box doesn't mean anything in the lowest bidder, mass factory produced world we live in. Unless you bought high end business class or server gear, it probably was not tested out of the factory. Stuff tested out of the factory usually has a QC receipt with someone's signature and a date. Even with high end, high availability gear, you should still test it. Pushing power and loading Windows is not testing the build. Bad parts aside, testing can also reveal novice build mistakes. If you are not a technician and don't know how to spot build mistakes, a complete teardown and re-check/reassembly might be worth your time. Once you eliminate bad parts or build mistakes, then you can worry about drivers, OS corruption, etc.

You can try to recalibrate the battery. This method will work for most Li-ion powered devices: Unplug the device from power source and power it on. Drain all power from the battery. Keep screen on, play games, etc until auto shutdown. Keep powering on the device after it auto shuts down from low battery until it will not power on anymore. Leave the device OFF and let it charge overnight. Power on the device. It should now read 100% battery and 0% will be more accurate. If this method does not work, replace the battery. This method does not magically fix an old battery, it simply let's the device learn the new 100% charged and 0% levels. All batteries will degrade over time and will need replacement if recalibration does not work. This method also does not give your battery more life. It can seem like it does because the smart charging will learn the new proper 0% level and not shut down so quickly. The real issue is your battery has degraded and only this method can tell you how much. You can expect about a 2-3 year lifecycle for Li-ion batteries before they need replacement. You would be surprised how often you can resurrect old phones with this method or simply replacing the battery.

Mikrotik has some low cost prosumer/business class router offerings. We have converted several businesses to Mikrotik that have lots of public wifi and private wireless users and it handles the load great. hAP AC comes to mind. It is a triple chain capable AP/router combo. hAP AC lite is cheaper, has 100 Mbit instead of Gigabit and less CPU/ram. It should be fine for your smaller home network, but doesn't give you as much room to grow like the non-lite version.

Good working PSUs don't make a funny noise. As a technician, I would put that PSU on my tester or at least use a voltage meter to test it. Alternative test would be to try a different PSU. Also disconnect that PSU from the entire system if testing. Blown PSU can send a surge through the rest of the system and take out other parts with it. If you are talking about an NVIDIA GTX 970, the minimum power supply recommended is 500W, not the 450W in your part list. If this is the case, you have likely overloaded the PSU. For good measure, scale up a few steps with the wattage. PSU efficiencies degrade over time. 500W recommended? Go for a 650W+. 600W recommended? Aim for 750W+

VNC is a popular Go-to app, but beware of the security implications of running unencrypted VNC. There are ways to encrypt VNC. RDP is much better on supported OS due to native UI handling, it feels snappier.

Have you tested the ram sticks? Booting into Windows is not a test, something like Memtest86+

Event Viewer is useful, but it doesn't tell you what is wrong. It's like going to a doctor and saying "My stomach hurts..." except in this case it looks like a million times saying "My stomach hurts..." Establish a working baseline of tests, then look at getting into software troubleshooting. Most DCOM stuff is really only useful to programmers that make those specific apps. Unless you are the programmer for that CLSID and APPID that is doing the complaining in Event Viewer, you are just looking at "My stomach hurts..." with that screenshot. Also, there are many ways to "reformat" and "reinstall" Windows 10. Built-in "refresh", CD/USB installation media, factory reset using OEM image. The only way to be 100% sure of a fresh installation, uncorrupted, is with separate installation media on known good working parts, not built-in. As a technician, "Just bought the computer parts yesterday, last week, month, year" is not a baseline established for a known good working system. I have seen PCs fail out of the box, never turned on.

Depends on the program. As you install each program, you should be able to choose the destination.

Did you test new parts after installation? And test again now? The joy of computing is that a million things all have to work perfectly and one can fail at any given moment without notice or symptom causing the entire system to suffer or crash. Cheap manufacturing processes means that only 1 out of XXXXX parts are tested before they are sent to stores/shelves/consumers. Brand new RAM installation? Test it. Just built the PC from new/unopened parts? Test it. Random crashing that doesn't make sense? Test it. Random crashing that does make sense? Test it. CPU/Cooling, RAM, and HDD for starters. Some of those "overpriced" parts and PCs out there are overpriced because that exact system and those exact parts were stress tested before you bought them. If you didn't reach for the top shelf models where the manufacturer uses strict testing practices and strict engineering tolerances, then it's up to us, the OEM builder / consumer, to make sure the parts work fine. Server stuff and workstation grade (business) stuff usually fall into this category of pre-tested equipment with high manufacturing standards. Most "gaming" parts fall into the consumer grade category, even "high end" brands and models made for gamers.