Overloke

Honestly, so was I - I have Intel in my current PC, my fiance's old PC, my media PC, and both my servers. I went with AMD for her latest build, and it's going well, so I thought I'd give it a swap. I'm also still running Windows 10, and unless there's been changes, I vaguely remember reading/hearing that Windows 10 doesn't quite know what to do with P-cores vs. E-cores. Prices go up and down, the total cost of this has fluctuated by about $200 over a few months as sales come and go. I've never used iCUE for my RAM. I don't think I even use it for my AIO anymore. Where are you getting your parts/prices? The 4080 I'm looking at is $1,699 AUD, and a 4070Ti is minimum $1,099 AUD. So sure, it's $600 more, and maybe I don't need the 4080 for a 2560x1080 ultrawide, but I'm sure I'd want to upgrade my monitors somewhere down the line as well, and I'd hate to be caught out because I run out of VRAM at higher resolutions. My main goal would be "Run Space Marine 2 when it comes out in September at absolute max settings." Thanks dude. This isn't my first PC building rodeo, but there is definitely still some tension every single time you press the go button for the first time!

Hello from Australia! At current, I have a venerable (for the speed at which technology progresses) gaming PC from around 2017. It has the following: MSI GTX 1080 Intel Core i7-7700K overclocked to 5GHz all-core Corsair H100i v2 240mm AIO 16GB of Corsair Vengeance DDR4 @ 3000MHz FSP Raider 650W PSU MSI Z270 Gaming Carbon X motherboard 2 x Crucial MX500 2TB SSD in RAID0 (I know, I know, but I run disk backups to my NAS, and I just found another one laying around, so why not?) 1 x 2560x1080 35" ultrawide @ 144Hz, and 1 x 1080p 27" @ 60Hz. Nothing special, they're not amazing 'gaming' monitors. This absolute gaming chad has served me well for years. However, it's getting a bit long in the tooth - 45 FPS on Helldivers 2, if you can believe it! And so, I believe it is time to give the old girl a new lick of technological paint. However, due to everything being from that era, and having not upgraded the system piecemeal as I went along, I now have to update everything at once. Here's my current plan. Give me your thoughts, let me know what you think, am I missing or overshooting anything? All prices are listed in Australian Dollarydoos. Motherboard: MSI MAG B650 TOMAHAWK - AUD$345 I don't want to get a bottom-of-the-barrel motherboard. Four RAM slots, two M.2 slots (although I hate M.2 drives and if you check my post history you'll see why), built-in Wi-Fi (although I'll be wired in), buncha SATA ports, etc, etc. If there's a better value-for-money motherboard, lemme know. CPU: AMD Ryzen 7 7700 - AUD$529 It's a 7th Gen Ryzen, what's not to love? Going up from 4c/8t to 8c/16t, higher boost clock, more cache, and I'm sure an absolute whack-ton of IPC improvements in the last seven years. I don't need mucho nutso performance (e.g. the X3D range or anything like that. I don't plan to stream or do rendering or anything). CPU Cooler: be quiet! Dark Rock Pro 5 - AUD$125 The seller I'm looking at doesn't have any good Noctua stuff in, and I've used be quiet! before at work, and they're lovely. It's possible the built-in Wraith cooler for the 7700 would be fine, but I do like my overclocks. Although from what I've heard, overclocking doesn't get you the performance boost it once used to. Also, I'm sick of water cooling. GPU: MSI RTX 4080 Gaming X Trio - AUD$1,699 The moment you've all be waiting for. The chunkiest boy. The biggest upgrade. I was looking at a 4070Ti initially, but people seemed to be saying that the VRAM could become a limitation on that card, and the 4080 is a few hundred dollars more, but worth the money. Thoughts? RAM: Corsair Vengeance 2 x 16GB DDR5 @ 6000MHz - AUD$169 It's dedodated wams, what more do you want from me? PSU: Thermaltake Toughpower GF A3 850W - AUD$149 Probably overkill, but my current 650W FSP Raider doesn't have the cables required to power the GPU (I believe). I prefer too much than not enough. If there are better quality power supplies, lemme know, but as long as it's name brand, it's probably fine, right? TOTAL COST: AUD$3,016 Note that this build obviously isn't set in stone, I haven't bought anything yet, and prices can still fluctuate. However, I think it's finally time I start thinking seriously about upgrading, and I don't want to half-ass it. Whole-ass only. Thanks lads and ladettes.

I bought a Chinese Huananzhi QD4-X99 motherboard and an Intel Xeon E5-2670v3 off of AliExpress for like $150 AUD, which is about 78 GBP. Obviously you run the risk of a sketchy seller on AliExpress, but you can see the sellers that have decent reviews. I got my bits and pieces like over a year ago and they're still running perfectly fine, no hassles at all. It's running about half a dozen VMs, including both a Minecraft and Terraria server in the past.

I have a Synology DS418 4-bay NAS. I'm just running it in a JBOD at the moment, until I can scrounge the funds to get some 8TB NAS drives and slap it into RAID5. Earlier tonight, I found a 3TB drive kicking around, so I deleted everything from the NAS, shut it down, and swapped out an older 1TB drive with the new 3TB one. When I turned the NAS back on, it appears that almost all of the applications I had on it don't work anymore, likely requiring me to reinstall and configure them. Not a HUGE pain, but certainly an annoyance I didn't foresee. I fully expected to need to nuke my storage pool and everything when swapping out the drive - since it's just a JBOD and doesn't have any redundancy - but I'm just curious if anyone else has run into the same issue where simply swapping out a drive as part of a JBOD pool resulted in a BUNCH of applications from the Package Center no longer working. Even weirder, there's apparently meant to be a 'Repair' button - that's not there. I also install the Logs Viewer, but that doesn't show any log relating to any application issue either. The note says to go to Storage Manager for detailed information...but there's nothing in Storage Manager to indicate any issues. This also persists through reboots. I REALLY don't want to set up DSM from scratch again, and I VAGUELY remember something like this happening YEARS ago when I last swapped out a drive. If anyone has any thoughts, that'd be cheers.

Personally (and I had a lot of pushback on this when I mentioned I had done this new build) you can get old Xeons and Chinese-branded X99 motherboards from AliExpress. I got a Xeon E5 2670-v3 (12c/24t, single-core boost to 3.1GHz) and a Huananzhi QD4 X99 motherboard for like $150 AUD - and the Xeon itself was like $12! I still have my original server, an mITX box running a Pentium and other consumer parts, but that didn't offer a lot in terms of virtualisation and things like that. A lot of people were saying that the motherboard would crap out, I'd experience crazy issues, going through warranty would be an absolute nightmare, etc, etc. And yeah, I'm sure that's a possibility, but it's not what I've experienced. It's been rock solid, running about five VMs pulling a variety of duties (webserver, Plex server, Tailscale subnet router endpoint, etc), and it was running modded Minecraft and Terraria servers simultaneously without too much grief. I suppose it depends on how many players you're expecting to have. Throughout all this, it's been running Windows Server 2016, and I know in terms of 'What's the best server OS', unless you have a specific use case (AD, etc), Windows is pretty far down the list. However, it's what I'm used to and it's been perfectly fine. Some of my VMs run terminal-only Ubuntu server though. If you absolutely need nutso-fast single-core performance - or if you just don't trust potentially sketchy Chinese hardware - then it's not for you, but it is a way to get decent, if old, server hardware that you can hide away somewhere and forget about.

Hah, no problem. I have a ton of them. As I said, I think the problem was very likely dirty power. One of the original 2.5" SSDs that was experiencing corruption problems was my first server SSD, and I never had a problem with that. Now the 1TB Silicon Power is in its place in the server, and I have no problem with that either. I don't think the Crucial one was low quality. We initially bought it for a customer, he ended up not needing it, it floated around the office in its enclosure for about two years, I took it home, removed it, and put it in my partner's desktop, and it's been fine for the last, I don't know, two months or so. I think you were spot on with "Something sketchy happened at time of boot", and I think that 'something sketchy' was probably 'the power cable was hanging really loose in the PSU socket and probably sent a bajillion micro power spikes in half a nanosecond'.

I don't remember what the make and model was of my first one. It was just a bad drive, pure luck of the draw. The second one (the original one I got for my partner's desktop), was a 1TB Silicon Power P34A80. It's now living in one of my servers that I'm not too concerned about. I got it because my mate said he had no problems with Silicon Power SSDs, conveniently not mentioning that he meant 2.5" SATA drives. The third one (the one that shit the bed tonight) is a 2TB Crucial P1 that was originally part of some Simplecom external M.2 adapter that I took from work. A power board, in the Australian terminology, is one of these: LINK - also known as a surge protector (although not all power boards have surge protection built in). The board I was using, as I said, looked about 20 years old and made zapping noises when I was first getting everything set up and plugged in. I have considered buying a UPS for our machines. However, space is at somewhat of a premium (even though I know they are small), and I HOPE the problem will be fixed just by cleaning up the connectivity we currently have, with the new board and cable. It definitely makes sense that it could've corrupted at time of power on instead of while it was off.

I've always had HORRIFIC luck with M.2 form-factor SSDs. My first one was DOA, the second one that I put in my partner's PC corrupted and lost its boot table (was able to recover data and format it), and now the THIRD M.2 I've ever used (pre-installed laptop M.2's excepted) has just corrupted its partition table tonight, apparently while the PC was off. Last night, my partner was using it as normal, no problems at all, playing games, on Netflix, the usual. Shuts it down over night. She goes to turn it on today and whoops, won't boot. It tries to run Windows Startup Recovery, and fails. I had an M.2 enclosure, so I put it in that and plugged it into my PC. The main data partition was marked as RAW, meaning it had corrupted. Both sfc /scannow and chkdsk failed - they couldn't even begin to run. Nothing had been done to the PC overnight, the case hadn't been pushed or knocked, and I'd definitely installed the SSD correctly when I was swapping out the second failed M.2. As I was pulling the PC apart, I noticed a few things: The power board that the PC was connected to looked about twenty years old. I'd also remembered previously that it made MAD sketchy noises when I was plugging things into it. The power cable running into the PSU was loose, on the PC side. I could pull it out with one hand without needing to steady the desktop with the other. Now, since the second M.2 fiasco, I've been taking weekly backups of my partner's PC to my NAS, and it's in the process of being restored now. If it works, I've made a promise to myself to buy my NAS all the nice things it deserves - no more JBOD made up of mismatched 3.5" and 2.5" drives! There's even an SSHD in the mix! I've also swapped out the power board and power cable with better ones, but we won't know how she runs until the restore finishes and I plug everything back in. The only thing I can think of happening is somehow the dodgy power caused issues over time. Maybe not even noticeable while the PC was running. Another reason I think this is the case is because my partner's OLD desktop - admittedly using 3rd- and 4th-hand 15-year-old parts, including SSDs - suffered almost identical SSD partition corruption problems, and they were 2.5" SATA drives. Does anyone else have any thoughts or ideas? If this M.2 gives up the ghost again after the restore, I think I'll just need to accept that I'm cursed by the M.2 Gods and be forever stuck with 2.5" SATA SSDs forevermore.

To be fair, most of the Crucial-brand SSDs I've gotten up until now have been the 2.5" SATA MX series, and they've been fine. I got the current one for free, from work, from an external SSD enclosure, so I'm not surprised it's not top-tier, honestly. Don't know what happened with the original one I bought - the Silicon Power P34A60 that had good reviews. I literally got the Crucial P2 for free lmao, and I've always had great experiences with Crucial in the past. Admittedly, with 2.5" drives. Anyway, since I've disabled D.O.C.P, there hasn't been a single issue, so I'm going to assume that was the problem, and I owe Crucial an apology since their SSD appears to be perfectly fine. I may try some manual RAM overclocking to see if I can get closer to the D.O.C.P spec, but honestly, the missus doesn't really care enough for it to matter right now.

Looks like the CPU (Ryzen 5500) also only technically supports up to 3200MHz RAM, so running it at 3600 could also potentially be causing issues, I suppose.

Brutal. Odd that running it in D.O.C.P spec is enough to kill it, but so be it. I probably won't look at getting a new SSD, the gf doesn't do enough to really see an improvement from the RAM speed increase anyway.

I don't know exactly, I didn't scope it out when I had it outside of the case. Based on some pictures I just looked up, it APPEARS to be a Crucial P2 2TB, model number CT2000P2SSD8. Disabling D.O.C.P appeared to fix it, so I might just spend a little bit of time manually tweaking things tonight. Now that I know it's (probably) not the SSD, I feel a bit safer. I've taken a Windows System Image from within Windows, and another one from EaseUS ToDo Backup as well, so worst case, we only lose a day or so, and it's a complete system backup.

As an update, I disabled D.O.C.P (which I had enabled mid-ish last week), and it appears to have fixed it for now. I found it odd that RAM issues would stop the SSD from showing, even in the BIOS, but I guess it's not impossible. I might need to do some manual RAM tweaking now, since the D.O.C.P speed was 3600MHz, and the 'Auto' speed is 2133MHz, which is a SIGNIFICANT drop, especially on an AMD system (which this is running).

tl;dr M.2s hate me, my gf's SSD is bluescreening, and the only thing I can think of it being is a dodgy power board, but other ideas are welcome. I'm going to preface this by saying I despise M.2 drives. I've tried three of them in desktop machines now, and each one has shit the bed in some form or another. The first one I put in my desktop was DOA, and after an entire weekend troubleshooting issues with it, I returned it under warranty and got my money back. Two years after that - earlier this year - I was building a new PC for my girlfriend. I thought 'That must have just been a once-off, lots of people I talk to swear by M.2 drives, and if they ALL crapped out, nobody would use them'. I grabbed a SiliconPower M.2 SSD for her machine, and it seems to have just nuked its own boot sector within six months, no longer booting and requiring internment into an M.2 external SSD case for me to pull data from it, where it currently resides. I was going to buy a 2.5" SATA SSD and be done with it, but I was able to snag a 2TB Crucial M.2 SSD from work. I've had great experience with 2.5" Crucial SSDs, so as a last hurrah, I put it into her machine, reinstalled Windows, and copied whatever data she wanted across from her old machine. This Friday just past, she told me her desktop had blue-screened, with the error WHEA_UNCORRECTABLE_ERROR, and rebooted into the BIOS. I poked around, and her SSD isn't showing in the boot menu. Oh no. I powered down the PC, and after about five seconds, turned it back on it. It booted into Windows. Whew, just a weird once-off. Earlier this morning, my girlfriend woke me up. "It's happened again". Same deal, same error, same no-show in the BIOS. I unplugged her machine and re-sat the SSD. It was installed correctly initially, but it's a simple thing to do. Powered up, booted to Windows. Excellent. Fifteen minutes later, "It's happened again." Same deal. I again just turned the machine off for five seconds, then turned it back on, and it's fine. Her machine has a quality SSD, a quality power supply that is over-specced for her system, and the SSD has been running fine for, ooh, about two to three months now. The ONLY thing I can think of at this point is that the power board her machine is plugged into is a bit scuffed. Power points are very difficult to plug into, it crackles sometimes, it's dirty and janky. HOWEVER, MY machine runs off a power board that connects into hers, and my machine has been rock solid - barring the initial M.2 SSD incident - for five years straight. I plan to replace the power board, since it'll only be like $20 for a decent one, but does anyone else have any thoughts as to what it could be?

I decided to learn about setting up websites and stuff, just for a laugh. I spun up a VM on a home server, loaded up Ubuntu Server, and spent, I dunno, two days researching, setting up, configuring and troubleshooting. I spent about four hours troubleshooting why I couldn't access my website from my local network (turns out that Tailscale, while a good service, freaks out if two machines are on the same network and it doesn't know what to do). I eventually also decided to poke around with SSL, and set up an Origin server certificate via Cloudflare. After more fricking around, it worked. Fantastic, I have a website at ABC.XYZ.link that is secured via SSL and I have the little padlock in my address bar. However, when I go to www.XYZ.link, or just even XYZ.link, it says the certificate isn't valid. I set up CNAMEs in Cloudflare to redirect XYZ.link and www.XYZ.link to ABC.XYZ.link, and the Origin Server SSL certificate I generated covers the main domain, plus a wildcard for subdomains (e.g. XYZ.link and *.XYZ.link). I'm using apache2 on Ubuntu Server 22.04, if that matters. If anyone has any idea why the certificate is only covering a single subdomain, that'd be cheers.