You need to understand that a GPU does not conjure details out of thin air. Higher settings and higher resolution also mean more CPU load. The difference in GPU load is usually larger, so you're more likely to be GPU bound, but it's not always the case. Source engine is especially bad about this, where way too much is done on the CPU.

So you complained about 4K High benchmarks and pulled out 4K and 1440p Very High benchmarks instead. Not impressed, I must say.

Let's go through the benchmarks in reverse order:
We're not going to play some mystery guessing game, so the LTT benchmarks are completely worthless.
As you can see, at 4K the 4090 driver is fucked, at 4K the 6950XT is the fastest and at 1440p it falls behind even the 3090. So the only apples to apples comparison are the 3090 Ti and 3090 and their difference is ... 39 fps, regardless of resolution and total fps. Logically the prediction would be that at 1080p Very Low we'd see something like 1000 fps vs 1039 fps. Not very impressive.

The 3080 vs 4090 is pretty scuffed because the average the 4090 shows 450-500fps average on 1080p low and 500-550 average on high, no idea how the current fps are so much higher than the average for the entire run, the 3080 shows 650-ish on low and 670-ish on high. So I'm not sure what're you trying to show.
Where could the difference be coming from? I don't know, maybe it's the fact that the 3080 was run with 4800 MHz RAM and the 4090 with 6200 MHz? CPU clock is also different.
No idea where you're getting 46% more fps from, but it's most definitely not an identical setup and not even the same benchmark.
Again, worthless.

Then you got a bunch of benchmarks showing that lower details/resolution lead to higher fps. No one is suprised by that. See right at the start, GPUs aren't magic devices that pull more pixels out of the aether. Higher details/resolution also mean more CPU load.

Which brings us the only benchmark that could've been relevant.
12900K + 3090 vs 12900K + 3090 Ti.
Except one is DDR5 6400 30-37-37-26; 12900K (SP103) / P Cores 5.5Ghz / E Cores 4.3Ghz / Cache 4.5Ghz
the other is DDR5 7200 30-41-40-28 12900K (SP103) / P Cores 5.7Ghz / E Cores off / Cache 5.2Ghz
And there's your answer.
CS:GO getting scheduled on E-Cores sucks and source engine loves fast RAM and Cache. That's all there is to it.

So you've got benchmarks that prove what you claim, but you don't want to show them because even a single screenshot would be too much.
Yeah, that's copium.

A better GPU doesn't help at all though if the game is already running into the CPU limited. This is not some cooperative effort, the GPU only starts working once the CPU is done. That's that mysterious "maximum pre-rendered frames" setting.
Higher clocks, more and stronger compute units, wider bus, faster memory, three other ways to say faster memory, they all don't matter. Whether the GPU needs 1ms or 0.5ms to render each frame does not matter when the CPU only send a new one every 2ms.

You're going to see a couple more fps with faster GPUs, but once you're really all the way at the CPU limit, no secret sauce in the GPU will make any difference. E.g. if a 1070 gets 480 fps and a 1080 490 fps and a 2080 495 fps and a 2080 Ti 496 fps then you can damn well be sure that a 4090 isn't going to get 550 fps but rather 499 fps.
Not sure what CS:GO benchmarks you've seen that contradict that.

In fact, the newer drivers are probably going to make things worse for a while.
Most don't bother benchmarking CS:GO anymore because there's nothing to learn from getting almost the same fps with half the cards you're testing and it screws with averages, so the only thing I've got to offer is this:

https://tech4gamers.com/wp-content/uploads/2022/10/RTX-4090-vs-RTX-3090-Ti-Test-in-9-Games-5-28-screenshot.png

Doesn't seem all that great, does it?

tl;dr
Are you going to see a noticeable difference between a 960 and a 4090? Yes.
Are you going to see a noticeable difference between a 1080 and a 4090? No.
The difference comes from having a card a bit faster than just about reaching the CPU limit, not from having one that's extra new, extra shiny, or 3 times faster than needed.

The point is that now the comparison of 5600X and 12400 actually makes sense because they both cost ~190.
So in terms of price the 5600X became the 5600 murkscribe wanted, while the 5500 has taken the place of cheapest recent 6 core CPU and the 5600 is somewhere in limbo between them.

Oh, I never said it's anywhere near the same budget, just that it's tempting.
And for all we know murkscribe is going to wait so long that the 7600 is released.

Oh yeah, forgot to mention that and the BIOS update.
The 5500 is basically a 5600G with the clockrates of a 5300G and without the iGPU, so same advice as with the 5600G applies.
A B450 mobo whose BIOS can be updated without a CPU will be cheaper at the cost of fewer USB ports and no PCIe 4.0, which doesn't matter in case of the 5500 and 5600G because they don't have it anyway.

The build I linked is a mix of things, I don't really recommend spending 50 quid on a cooler for a CPU that only costs 120.
Basically that cooler and mobo and 5600(X) for overclocking, B450 mobo with BIOS flashback and no/stock cooler for a 5500 or maybe 5600G.

To be fair, a 5600X now costs less than 200$, the MSRP of the 5600. And the 7600X is very tempting.

#3882
The 5600X is a bit too old to be the best deal at this point. There's a reason why the 5500 and 5600 were released this year, they simply fill the budget niche much better due to being much cheaper while not being that much slower.
Especially an overclocked 5500 is really hard to beat.
The 5600G is technically slightly faster than the 5600 while including an integrated GPU and being slightly cheaper, but it doesn't support PCIe 4.0 so it's a bit of a sidegrade. Worth considering in combination with a B450 mobo to get 5600 performance at 5500 price.
A cheaper cooler should also suffice for any of these, even for overclocking.

Though honestly it boils down to this: If it's about the budget the 5500 with the stock cooler is unbeatable, if it's about performance then the 5600X and 12400(F) might be very close but 12500 and 12600 also exist.

Other than that cheaper mobos exist, a 750W PSU is a bit overkill for a 400W build and a cheaper/smaller case would be possible.
https://uk.pcpartpicker.com/list/9yK6PX

#3878
The 5600 is a 150$ CPU, the 7600X is a 300$ CPU. Are you asking me if a 300$ CPU is worth it?
It is ~35% faster, no idea if that's worth it to you.

For lower tiered 7000s you'd have to be fairly patient. The 5600 was released 2 years after 5600X because AMD was doing far too well selling the more expensive models to bother with cheaper ones.
APUs are planned for the end of 2023, so that's still a year away.
You could wait until Intel releases the 13600K and so on (20th October) and hope that reduces the 7600X slightly, but that's about it.

#3879
This is about CPUs. No, using the exact same numbers for both is not doing AMD any favours.

#3880
Not a masters but yes.
Though this is a hobby and has little to do with computer architecture.
Basically, this page scratches the surface of computer architecture, the other 20 tell you if it's worth buying.

Again, why does that mean you need to import an XG2431?
You won't get 250 fps in hunt, will you stop playing that?
What is stopping you from buying a monitor available in the UK that can do strobing at any refresh rate?

You need to decide either on a budget or a performance goal. You can't both be budget conscious and then blow an extra 200 quid on a PSU you won't ever fully utilize unless you spend 2000 on a 600W 4090 Ti (if that happens and it's even that "cheap") and another 2000 or so on a 300W 32 core Threadripper.
I'm not sure why you expect people to buy a new CPU every year. Or is there some special reason why some brits would want to get rid of their cpus early next year?

I don't think 250 fps in hunt are happening, the rest should be easy.
Not that 250 fps on a 240 Hz FreeSync monitor are going to do anything. Unless you want to import a ViewSonic XG2431 specifically because it can't do strobing and variable refresh rate at the same time? That's certainly one way to make your life harder.

Your plan is still shit though.
Do you really think you'll get a used 5800X3D 5 months after release?
What is the point of a "higher end cpu" if you're also going to handicap it by going with DDR4 RAM that was considered good in 2017?
If you want to save money, have you considered not spending as much on your PSU as on CPU and mobo combined? What's a 1000W PSU supposed to do for a 500W build?

There was no need to bring this up during i69.

We did just fine not knowing about this video for two weeks, I'm fine with not letting it affect the next two months or years either.

There is little point in mining with a card as old as the 9500 GT. The electricity would cost much more than the mined coins are worth. If you just want to transform money into heat you might as well burn a stack of one dollar bills.

The GTX 680s aren't going to be much better.

Mining doesn't use SLI. Any PCIe slot would do, even x1.
I've got no idea why you're looking at a LGA2066/X299 mobo, you don't have a CPU for that.

Sometimes, the reason why everyone is telling you that it's a stupid idea is simply the fact that it is indeed a stupid idea.

Just go with the secondary gaming pc. Forget about the 9500 GT, get some used DDR3 RAM and a mobo, ideally Z97 because overclocking would be nice, a PSU that won't set itself or anything else on fire, and a case recent enough to fit the mobo, optionally even recent enough that just looking at the layout won't make you want to kill yourself, and you'll have something actually useful.

Make sense for what?
We're also 1.5 months away from the Ryzen 7000 release, so if you're willing to wait until 2023 for an RTX 4080 or something like that then why bother planning anything out now?
Also ngl, 2400MHz RAM would suck for Zen.

Always disappointing, but not surprising to see that things really haven't improved at all since 2015.
Thanks for reminding me about memory latency. TF2 does seem to be affected by that significantly, so it's possible that 5800X3D might end up being the fastest CPU for TF2 due to cheating a bit with its massive cache.

#3860

That's usually power failure or anything that could cause a hard reset, I think.
Check all cables, especially GPU power, check RAM (whether it's seated properly, maybe even run memtest86+), update BIOS and drivers, all that fun stuff.

#3861
Really bad.
Back when I check years ago, the answer was theoretically infinite, so it will use all cores/threads for the parts that are multithreaded, but so few of it is multithreaded that the average works out to less than 2. Plus one more for the GPU driver, so a third thread helped, but not a lot. As in, 2 cores were significantly slower, but 2 cores + SMT/Hyperthreading for 4 threads total was only marginally slower than 3 or 4 real cores. And beyond that nothing really changed.
The last time I asked, it seems to have gotten slightly better or just worked differently for Ryzen since those seem to be getting more frames beyond 4 cores/threads, but you definitely still shouldn't buy a CPU for TF2 based on cores.
You want clockrate/IPC and anything that is actually fast that way and/or overclockable will most likely have more cores/threads than you benefit from in TF2.

The game you're thinking of is Overwatch.