The simple goal: to build a sub – $1000 PC that will not be completely obsolete in a year or two.
My main PC uses: gaming, video, music, surfing. In that order of importance.
Not a very difficult project, all things considered; you can put together a decent $500 rig that will allow you to play pretty much any game released to date, and stay reasonably competitive for about six months or so.
You may have to make some concessions on resolution and framerate for the more demanding games (the good old ‘will it play Crysis?’ is still a worthwhile question, no matter how tired of hearing it you may be). You’ll definitely be leaning more toward last-gen technology, goosed with judicious overclocking, to get the most bang for your buck.
My somewhat more-demanding requirements make this sort of build less desirable. I need something that will withstand a dusty and hot climate, and be a bit forgiving to extra manhandling. I also need something that will still be a serious (if not necessarily major) player in a year or two, and hopefully more.
Just to put it into perspective, my last main computer gave me four years of competitive performance before it was more or less obsolete, and then three more years of doing pretty much anything that I needed it to do, as long as it didn’t require state-of-the-art technology.
It has been observed that the ‘sweet spot’ for budget versus performance tends to be in the $700 – $800 range, and that’s what I’m aiming for. The extra $200 – $300 buffer gives me plenty of wiggle room for the most important bits (very likely the graphics card(s), but we’ll see…).
Any other requirements? Well, it will be sitting in a prominent place in my living room, so I’d like it to look decent… but that’s a minor consideration.
Should I accomplish my lofty goals then I might be able to go at least some amount of time without worrying about going shopping for a brand new version of something I just purchased… It really does seem now like you have to go out every eighteen months to purchase an updated more advanced version which is slowly but surely putting your older version right out of service. Sure, you may have a little time. It’s likely yours will keep working (at a reduced speed) for some time to come. Eventually, you can be assured that the newer version makes it so the older version goes obsolete.
If I am successful, I will be the first person in a long time to not really have to worry about that sort of stuff. At that point I will count my lucky stars, not until. Plus, I am keeping notes on the whole project so I should be able to replicate it without to much effort should the need arise.