Admiral Regulus

Yep. I log in here like maybe once every 3-4 months. I know the feeling.

I've played the first two. Haven't played the third because I've been boycotting EA's Origin service.

Thank you!

I've never worked with a Phanteks case before, but they look pretty nice. That one looks like it would be a solid choice for most builds.

With the RTX cards, I'm a bit torn on the whole ray tracing thing. On one hand, I see it as a complete gimmick that should be avoided by any value-seeking consumer. On the other hand, I understand what ray tracing is and I know it's the future of rendering 3D graphics. Needless to say, I'm actually quite impressed by what Nvidia has done, and their AI denoising algorithm is very, very clever to say the least. It's the future of gaming for sure, it's just that the future is still kinda far away, and the technology is still in its infancy.

Overall, I have a bit of distaste for the state of the GPU market right now. The GTX 1000 series was awesome, but I hated how expensive it was without any competition from AMD. The GTX 2000 series takes that a step further. The performance isn't breathtaking, but it is a solid step up from the 1000 series--with, of course, a solid step up in price.

It's totally off the radar for me, as I'm happy enough with my GTX 1060 and can't afford to upgrade anyway. But for those who have money they're willing to burn, I wouldn't fault them for opting for a 2080 over a 1080 ti. There are architectural advancements made with Turing, and although it doesn't result in increased FPS at 1080p, it's clearly the card to have for upcoming AAA games and/or higher resolutions. 

This is especially true when we can count on GameWorks to ensure new games run like crap on the previous generation's cards. They've done it in the past, and you know it's going to happen again. That's why I dislike Nvidia.

Nvidia: great technology, horrible ethics. That's my view in a nutshell.

Come on, Derpy! You can do it!

Oooh! I know this one. It's 910.

It is.

What's wrong with it?

I don't intend to do anything today, and so far I'm right on schedule.

What are you researching?

That looks delicious.

I would add a correction in that you probably mean apoapsis, but I don't care enough to do so.

Wat.

That seems rather nice. Too much internet can be a bad thing.

Nice to see you again. What have you been up to lately?

Yep yep. One thing I've learned in recent years is that the "bad" people in the world aren't necessarily the ones I disagree with. Instead, the "bad" people are those that adopt this mentality rather than attempt to look deeper into the issues which divide us.

Wow. Yeah. That is really, really good.It kind of reminds me of a meme I saw a month or two ago.

I think the problem is that we tend to get into a routine and we isolate ourselves within that. So, we look within our daily life for something new, something welcoming, something to provide a source of happiness. The quick way to get that is to dive into some fictional world. But of course, it's not the real thing.

The best thing is to connect with real people. No one knows for sure what's real and what's not--that's not what this is about, anyway. There's a certain happiness you can get from being with friends and sharing meaningful experiences with others, and no fictional world is quite close enough to being an adequate substitute.

No idea. Mine are.

@Cwanky That's really neat. I hope you have a great birthday in 43 years!

I can't think of a better build at the moment.

I'm currently using an overclocked 1700 at 3.8 GHz and 1.35V. With the Fractal Design S36 AIO radiator I'm using, it stays around 70-80C under a prime95 stress test on all cores. Under day-to day usage, even gaming, it's around 30-40C even with the radiator fans barely running.

I love the 8 core Ryzens, because they're hardly ever fully utilized in day-to-day use. There's something satisfying about getting great performance, and seeing that CPU load is only at 20% at the same time. The only reason to buy a Skylake or Coffee lake is if you need 144 fps or more at 1080p or lower resolutions.

I would suggest perhaps waiting for the 1180 instead of getting the 1080ti, but considering the state of the GPU market, there's probably no reason to. With no competition from AMD, Nvidia is becoming the new Intel, and I believe we'll see the 1100 series cards as overpriced and only minimally competitive.

As for the case, I got a Fractal Design Define S for my own build, and I couldn't recommend it enough. It doesn't have any space for an optical drive, but there's a slightly different model, the Define R5, that does. You can watch a really good video review of it here.

I've built a PC with a low-end NZXT case before, and I liked it. Even though that one was cheap, it was pretty well thought out. It was nothing special though. With Corsair's cases, I find them to be overpriced and mediocre quality. I wouldn't recommend one at all unless you just really like their style.

Matlab, lol. I was using that just because indexing matrices is simple and easy.

I tried. Not very well, though. The code isn't well written so it's pretty difficult to trace.

There is actually only one possible solution (or, rather, there should be only one), but finding it is not an exact process with only the numbers given initially.

I think I came up with an algorithm that could work, but it's producing bugs at the moment. I have no idea why.

This is how it should work:

Assume 1/9 probability for each number for each space.

For each given number, go down the row, column, and block and set probabilities of that number occurring to zero. Likewise, increase probabilities of all other numbers accordingly (e.g., if a spot can't hold a 7, 3, or a 2, then the probability of all other numbers increases from 1/9 to 1/6).

Once that is completed, find the space that has the fewest possible numbers. Ideally, there is only one possible number here, but there may be two or three possible of the nine.

If there is one possible number, fill in the space with that number. Otherwise, create a copy of the puzzle, one for each of the possibilities. Continue with each copy until a solution is found.

It isn't a linear algebra problem.

Even the easy Sudoku puzzles don't give you enough information to solve for a solution directly. There is some amount of trial-and-error involved. So, my idea involves creating an algorithm which determines what numbers could fill each space (as it's never just one one number for one space), and then instead of randomly guessing, it would determine the probability of each possible number and use the most likely result as a guess.

Nice. I'm sure that now that you've done it once, you'll probably never want to buy a prebuilt again.

What kind of parts did you decide to go with?