cynexit

I would argue the values are correct then, but if the status is still "Good" you do not have anything to worry about. It has often been reported that SSDs can take way more writes than the manufacturer says on the fact sheet (see here and the end of the test). As long as there is sufficient free space (15%~20%) on the disk to trim and shuffle blocks without issues then there is a good chance drives can take much more TB written than expected. That being said: I would keep Samsung magician (totally agree on the UI btw) running in the background so it can alert you if the drive is actually starting to fail (afaik it should then go into read-only mode).

Some manufacturers don't really stick to SMART standards, I had this with WD drives which basically required you to use the SMART monitoring program from WD to get correct readings. For this reason I would say give samsungs own SMART reader a go, the values maybe make more sense in there.

zoom can do at least 20 people in grid view (maybe even 25, don't know anymore).

As written in the top post: Anything goes as long as it supports a full-size ATX board. This is not for a practical build or anything so no mATX or ITX.

Hi everyone, I am currently thinking about a modding project and looking for the smallest "normally available" case that has support for a full ATX motherboard (and possible GPU). So far I especially found the Thermaltake Core G3, but I am sure that can't be the smallest one considering how creative some of the really small ITX cases are. Has anyone any other super small full ATX case in mind?

Yeah, a lot of people do that. If the external drive is an SSD and the port is USB 3 or better than you won't even feel a difference in loading times. The only thing to look out for is that the external drive is always assigned the same drive letter so that programs like Steam or Origin don't get confused.

That's like super not helpful, but thanks anyway ? So to recompile: https://www.awd-it.co.uk/ https://www.cclonline.com/ https://www.scan.co.uk/ https://www.overclockers.co.uk/ https://www.ebuyer.com/ https://www.box.co.uk/ https://www.alzashop.com/ https://www.aria.co.uk/ https://www.novatech.co.uk/ Thanks everyone for contributing! If anyone knows about anymore specialized shops like pexonpcs.co.uk keep them coming!

Good morning forum, I soon have to move to the UK for work reasons and wanted to ask what shops you guys could recommend, online and offline, for computer hardware. If you could attach some properties to the shop (e.g. really cheap, good customer support, quick exchange during warranty, etc.) that would be awesome too! Cheers, cynexit

As long as you don't need to compile Chrome from source you won't need a big CPU/much RAM for coding and compiling. Since you already mentioned "full-stack" I assume you do mostly web stuff, so tops you need a local apache/nginx + database and something like NPM/Composer/Go toolkit/cargo benefits much more from an SSD than from many cores. If you code mostly C/Rust/Go or any other language that needs to be compiled: as long as it's not a really, really big project, CPU still doesn't matter much, 4 or 6 cores won't make much of a difference, today's compilers are really optimized. For serious ML none of these GPUs is viable. If you just want to play around go for the 1060. All together I would opt for option 2 or wait and grow the budget, depending how serious your needs are.

I'm 6'0" too, so let me help you out here: https://www.sockdreams.com/socks/thigh-high-otk/thigh-highs.html This shop was one of the best discoveries of my life.

Just cutting power to a powered off system should not damage it whatsoever. Seems like your PSU is defective, if there is still warranty, try to get it replaced.

Alright, then depending on how tech savvy you are: 1) Try to log the traffic on your computer with WireShark and check the packets directly before and after the disconnect. 2) Use "Microsoft Network Monitor" (if that is still around, haven't used that in a loooong time [EDIT: Seems like it: https://www.microsoft.com/en-us/download/4865]) and check the logs there.

If this is your own residential connection, the log of your router should be a good place to start. See if the connection drops on your whole house or just your PC. The spike doesn't necessarily have to do with loosing the connection, it could also be that Windows tries to recover the just lost connection.

Without real benchmarks there is no way to tell for certain, but: games don't often scale linearly with CUDA cores, but rendering workload mostly does. So looking at the 2080 with 2944 CCs and the Ti with 4352 CCs one could (very very roughly) estimate that the Ti will be ~40% faster than the non Ti version. So now it fully depends on the price: if the Ti is less than 40% more in price you theoretically would have a net win in performance, if it is more expensive it would make sense to go with a 2080. But again as a final disclaimer: This is so much guesswork that you shouldn't give a crap about the numbers above!

Technically this could be used to speed up renders. But here's the deal: The render engines need to be updated to support it. I know some already announced that they are working on supporting it, but others have yet to mention it. My advice: Check the news/forums of the software you are using, find out if they plan to support the new hardware RTX brings and then decide from there. IF dedicated support is promised I would honestly go for it, if not I would just buy two older, by then cheaper cards and run them in SLI which should have more "raw horsepower" than one new RTX card. (I know it's technically not SLI in this context but you get the gist)

Soooo, your plan is to just swap in the aluminum block once you purchased it? You do understand that there is a good chance your loop will already be damaged beyond repair at this point, so even if you swap this one copper part, particles are in the system and happily eating away practically your whole loop.

Side note: The uncached website should be available without HTTPS because that can interfere.

This could be a really cool and small machine to use for ssh'ing into your servers on the go. Slam your favorite distro with a tiling window manager on it and you wouldn't even need the integrated track pad anymore.

Why though? The 980 ti is more equivalent to a good 1070/bad 1080 if you have a lucky card.

In Germany they go for 200 to 300 Euro, 350 if the seller is really lucky and the buyer is not very tech savvy.

Totally agree with you here. If I can have better temps AND quieter operation then fuck looks. On a side note: did you look at the Morpheus II from Raijintek too? Curious why you went with the Arctic.

I also started with VB6, looking back it is horrible, but back then it was pure magic. EDIT: Just read the OP, so obvious disclaimer: DON'T LEARN VB6! Start with something easy like Java, C# or Go and then try something like Scala or Rust to learn new, more advanced concepts.

I would be really surprised if someone released a laptop that can sustain that 4.6 ghz boost for more than a minute though....

Since I don't currently own a 7XX card I don't know anything for sure, but I remembered there was a LTT video about that and at least in their instance they didn't find anything suspicious: Mind you of course that this is again a sample size of one. I once had a similar problem with a GTX 285 back in the day. After a lot of searching (and finally buying a 770) I found the issue during dissasembly of the card. One of the heat sinks that was glued onto a row of VRAM modules did get loose and was just hanging in the big GPU heat sinks, so whenever the VRAM on the card was taxed harder (and reached certain temperatures) my machine would just crash and reboot. Of course temperatures looked normal on the GPU side because the die itself was cooled just fine, but keep in mind there are no sensors on the VRAM/caps/voltage regulators so there is no real way to find out about their well being. If you have nothing to loose and nothing else left to try I would recommend you to try and cool all other components besides the die better. Take the card apart, make sure all heat sinks and other cooling features are correct and then fix the fan speed of the card at 100% and rerun your games at high settings. If the card suddenly becomes stable it might just be that some parts aren't cooled as well as they used to be.

Might be the VRMs (area) on the motherboard. I assume you are using a CPU block and not a mono block so the chance is pretty good that there isn't enough airflow around the socket. VRMs can survive these temps but it is severely degrading their lifespan. Try to position a fan over the VRMs so that it blows directly down on them and repeat the gaming session, log the temps and see if TMPIN5 is changing. EDIT: Similar to this: https://www.overclockers.com/forums/showthread.php/720295-Custom-VRM-VCORE-and-Motherboard-Cooling-is-it-worth-it Just use a 120 fan and add some distance, you can try to "hang" it from the top or just use enough tape etc., get creative.