I really liked your videos on PC builds and I've seen your recent videos on workstation and console PCs. Any plans on making builds that focus not on a initial purchase price but longevity of the parts and technologies involved over a long period of time, ie. easy to maintain, reliable PC?
You've touched on builds for people who are supposedly "scared of PCs" but I don't believe you've yet touched on builds that are safe for people who have the "any PC I build isn't going to last more then 1 year before I have to throw everything in it to the garbage and start from scratch" stigma/fobia. Obviously the stigma has some truth to it, since if you skimp on every part and get the absolute bare minimum, upgrading (such as just adding something like a card to help with streaming) is somewhat of a mini-nightmare. It doesn't help that everyone suggests only the bare minimum on everything every single build guide you see; giving credibility to the fobia.
As an added question, if you don't have plans to make such videos on cost-effectiveness and easy to maintain over a long period, I was hopping if you could just tangible put a finger on what you would go for just off the top of your head? For example I know you like the 8350 very much, but I'm not sure how you feel about the other offerings from AMD, so if say you didn't like the "next in line" for upgrade from AMD how would you feel about the 8350 (and an AM3+ mobo) as the initial investment when you "had to" change it for a better one later?
Addendum
Here is an example of some basic guidlines, in case the context of the first question wasn't clear by itself:
initial cost doesn't matter; cumulative cost over the estimated lifespan of the build matters (lower is better)
lifespan is not set in stone (larger is better however); lifespan should last until the estimated failing point of one of the more critical components (eg. PSU, Mobo, etc)
lifespan includes the period where even after upgrades the central components (we'll take the mobo for the sake of the example) the PC in question is really only good for playing non-AAA and running OS + normal programs (ie. old but not a complete piece of junk)
safe overclock is okey (overclock to the point it might go into meltdown isn't; since it defeats the point of the build)
lifespan example: assuming 7 years component failure, 1 year AAA max settings, 3 years AAA max with upgrades (including overlocking), 4-5 years on playable settings, 6-7 non-AAA, but still usable for general work and indie/small games (ie. not at the complete bottom of the technology curve)
the following parts should not change from the initial investment: case, mobo, psu, ssd/hdd (adding more is fine)
every other part is fair game for upgrades; every upgrade obviously adds cost
resale value (on parts) shouldn't count since if it goes okey its a bonus, but the point of the pc is to squeeze as much as possible over a long period so the parts may be too old between upgrades
Again just an example.
I think there are a lot of people looking for builds with extraordinary longevity and/or simplicity, even if they don't necessarily come though the wire with all the uber tech savvy people talking.
Component failure is impossible to predict, you can pay 50 bucks on a budget board and it last 8 years or get a 300 dollar board and it comes in ready to RMA. The main issues a user can control is heat and power. Keeping components cool generally keeps them happy. A power supply with the smoothest energy deliverance also helps. Go for an I7 with a higher end board and extremely nice psu. Don't skimp on heatsinks or fans. The only part that will not last 7 years will be the gpu. Only 4 years ago was the AMD 5xxx series released, granted it would still handle low or medium on stuff you're better off buying mid grade and replacing every 3 years ish when it comes to gpus.
As tmlhalo said, you can't really predict failures in hardware, all you can do is do your absolute best to provide the parts with a comfortable environment.
If your aiming for 7 years of use, your going to have a pretty large outlay to start with.
First you want a case that is specifically designed with cooling in mind (probably air cooling), with a fair few fans and lots of dust filters and cable management options.
With the motherboard your gunna want something thats stable, personally id aim for top mid range boards, i find top end boards have a tendency to be a little unreliable in the long term.
CPU wise, if it were a build today id suggest an 8320 purely for its 8 cores, and cooling wise id get a really nice air cooler.
if your wanting 7 years out of a PC don't overclock end of.
GPU your going to want something thats pretty powerful, i wouldnt say get top end id drop down a card or two and leave yourself open to drop in a second card 2/3 years down the road.
PSU you wanna go top end here, probably looking at something silver/gold rated (maybe plat if you got loads of cash).
The thing that will make a huge diff in lifespan tho will be the user, simple things such as cleaning filters and heatsinks on a regular basis can add months to lifespans, simple stuff like removing the CPU heatsink for a cleanup and re application of thermal paste every 6 months or so helps massively. Keeping the system clean, by removing unnecessary files/folders and keeping a clean registry help out too.
IMO if longevity is your goal, you should probably ignore flashy new high end parts and look for the slightly older work horse tested parts .
I am going to have to disagree with you both. Component failure rate can be to some degree estimated. If you don't believe me ask me how many Dell computers I've had to repair for end users just days after the 1 year warranty. Be my guest look up the legal action taken against them. If you build with inferior parts you will yield bad results.
Buy parts made with quality components and good practices. You will reap the benefits, end of story.
Yeah i agree if your looking at pre built/ branded pc's that their hardware is built with a time frame in mind, generally individual hardware components don't suffer from this "greed" angle.
Im not saying they are all going to work 100% forever but they dont suffer from the light bulb conspiracy.
With regard to the 7 year figure, that's just an estimate, and this is a general question so the motive really varies wildly on a per use basis, so it really doesn't matter. I gave 7 years since it seemed reasonable and an average on what I feel most people would aim for and gave a reason for the sake of there being a reason. I feel you can legitimately base it on anything else (and people likely will) such as "technology curve," "time I care about the main components," etc, heck you can just roll a D20 +3 for the "lifespan." It really doesn't matter.
The heart of the question is: if you build for longevity what do you do differently as opposed to building for "cheapest buy NOW," and what saves you money long-term as opposed to saving money short-term. Both of you already gave some good suggestions, I'm just curious what the guys think, since I already have some idea what they think of the short-term variant.
I think you answered your own question at the beginning of the 2nd paragraph. You don't buy the 5 dollar cpu off the dude in the trench coat on the corner. Top of the line will last longer performance wise. That doesn't mean buy the most expensive everything. As also Namix have pointed out, not every dollar means an increase in quality. Some items have a lot of bells and whistles attached so the company can justify the price tag. GPUs on the other hand you're better off with the middle of the rode and replacing it every ~3 years. At the top of the line the premium per dollar is insane. Exp the Titan is 1,000 the 780 is 600. The increase in performance is ~5%.
Given the question of "cheap and now" has been answered pretty comprehensively, resuming the answer to this as "don't buy the 5 dollar cpu off the dude in the trench coat on the corner" I don't think is fair since it's an equally complex problem (until proven otherwise).
Understandably any answer would be generic to some extent since there's a lot of speculative science that needs to happen to answer it, but that's why I posted it to the guys here. I believe any answer they give would easily reflect off of the standard approach they've already used in their videos (their base as it were that we all have an idea of) and illustrate what key differences in build an average consumer with that mentality should look for in a more long lasting build; even if they only answer the problem succinctly.
Also obviously I have my own ideas of how to answer this problem, but I'm curious what Logan and the others think is the right approach. It's a question, but I'm looking for their opinion not any one answer.