Buying or building a gaming PC

Myranda

Double post

digitalscream

Myranda said:

digitalscream said:

speshul91 said:

On the subject of psu's if you've got a case you can see inside modular is the way to go to keep it looking tidy. 

There's a much better reason for keeping it tidy - airflow. Messy cabling screws up the airflow and almost always results in higher ambient temperatures.

Recently disproved by Linus Tech Tips... they actually packed sweaters into the case to totally kill airflow and still didn't see scary temp rises. Seems modern pcs just don't need the amount of air flow we all assumed.

Best reason for tidy cables (other than aesthetics) is for when you want to upgrade anything you're not wading through a jungle of cables every time.

Simple physics proves that their methodology (as you've described it, at least) is wrong or their conclusion is wrong. Try pumping roughly 200W of energy into a closed box and tell me whether the heat rises or not. Blocking the vents in a case is equivalent to that and guaranteed to increase the temperature relative to having a totally obstruction-free case. Given that it's a continuum effect...draw your own conclusion.

Of course, it also depends on how your heat is transmitted away from the main heat-generating components - if you're using heatpipes or water cooling to directly get the heat from the CPU to the back of the case, then the effect is obviously reduced. However, block airflow to the radiator fans by even 50% - like a mass of cables can - and you'll find the machine overheating and shutting down.

Myranda

digitalscream said:

Myranda said:

digitalscream said:

speshul91 said:

On the subject of psu's if you've got a case you can see inside modular is the way to go to keep it looking tidy. 

There's a much better reason for keeping it tidy - airflow. Messy cabling screws up the airflow and almost always results in higher ambient temperatures.

Recently disproved by Linus Tech Tips... they actually packed sweaters into the case to totally kill airflow and still didn't see scary temp rises. Seems modern pcs just don't need the amount of air flow we all assumed.

Best reason for tidy cables (other than aesthetics) is for when you want to upgrade anything you're not wading through a jungle of cables every time.

Simple physics proves that their methodology (as you've described it, at least) is wrong or their conclusion is wrong. Try pumping roughly 200W of energy into a closed box and tell me whether the heat rises or not. Blocking the vents in a case is equivalent to that and guaranteed to increase the temperature relative to having a totally obstruction-free case. Given that it's a continuum effect...draw your own conclusion.

Of course, it also depends on how your heat is transmitted away from the main heat-generating components - if you're using heatpipes or water cooling to directly get the heat from the CPU to the back of the case, then the effect is obviously reduced. However, block airflow to the radiator fans by even 50% - like a mass of cables can - and you'll find the machine overheating and shutting down.

The reason they ended up shoving clothing in was that bad cable management - and shoving things in the case wasn't having a significant effect... 

Cables just don't block airflow enough. Hell, cables, some boxes and some t-shirts (remember it being sweaters.. nevermind) don't take it into thermal throttle zones while running Furmark and Aida64 (75 degrees is warm but not a threat to the CPU)... back in the day of big thick ribbon cables maybe it was more of an effect, with already worse airflow, and hotter components... 

Sporky

I think there may be a mid-ground between "laminar airflow is essential for cooling", and "airflow is for losers".

Cold air in at the nottom, hot air out at the top. Job's a good'un.

digitalscream

@Myranda - the problem with that video is that they didn't really test real-world conditions. For sure, they've debunked the idea that airflow matters when you run your computers for 10 minutes at a time...that's not news.

For a start, they didn't test the ambient temperature inside the case. None of their tests lasted more than 10 minutes, which isn't realistic - ambient temperatures are a slow creep, and have far more of an effect over two hours+ (in the average gaming session, where the whole machine's being ragged to death). I also didn't spot any point in the video where they mention the temperature in the room.

Then you've got to contend with the fact that the average tower is not sitting on top of a desk, it's down by your feet and shoved in a corner. Add in a non-modular power supply and a much smaller case with drives cramping everything so that cables become a genuine impediment to air flow and coat them in dust, on a summer's day in a non-air-conditioned room...suddenly it really is a problem.

It's no single thing - lots of tiny little increases which you'd happily ignore in isolation. Thing is, they don't have to increase the ambient temperature much; they just have to get the internal temperature of the case up to the tipping point where the CPU and GPU coolers can no longer effectively get rid of the incoming heat. That can be as low as 32 degrees in some situations (yep, been there).

TimmyO

Any time I think I might like a Gaming PC (it happens every now and again) I read articles about building a good gaming PC and see recommendations for £600 (!) graphics cards and I run a mile. I end up thinking that if I'm not reliably going to be able to run at the highest settings then I'll stick to console.

I'm sure that there are 'better than console' results achievable with less than a £600 GPU but then it all seems a bit hot and miss :-/ 

Sporky

A £600 GPU ought to run new games at 4K with very high settings.

If you're playing at full HD then even half that might be more than you need.

The core of my gaming/general use PC is about 4 years old. All it's had is SSDs and two replacement graphics cards, and it'll run Fallout 4 at 4K quite happily. If I want to carry on with new games at 4K then it'll probably need a new CPU, mobo and RAM at some point, but not for a while.

This isn't a bad guide to how much to spend (and on what) based on the resolutions you want to play at.

http://www.tomshardware.com/reviews/best-gpus,4380.html

Bidley

TimmyO said:

Any time I think I might like a Gaming PC (it happens every now and again) I read articles about building a good gaming PC and see recommendations for £600 (!) graphics cards and I run a mile. I end up thinking that if I'm not reliably going to be able to run at the highest settings then I'll stick to console.

I'm sure that there are 'better than console' results achievable with less than a £600 GPU but then it all seems a bit hot and miss :-/ 

It's well known that the PS4/Xbone hardware is something like 5+ years old in terms of technology. You don't have to buy the latest and best GPU to compete with that.

Once you start gaming at 60fps in 1080p, you'll wonder how you managed before (PS4 is 30fps at 720).

For those thinking about building a PC, really there's nothing to worry about. It's like Lego these days, nearly everything only goes in one way!

speshul91

My graphics card cost me 130 brand new its a geforce 750ti oc. It runs most of my games on high with no issues at all. 

BloodEagle

I've got a GeForce 1070 -  costs about £400 and plays pretty much everything you can chuck at it 60fps 1080p+ . Can't stand playing at 30fps on consoles it gives me a headache

Myranda

digitalscream said:

@Myranda - the problem with that video is that they didn't really test real-world conditions. For sure, they've debunked the idea that airflow matters when you run your computers for 10 minutes at a time...that's not news.

For a start, they didn't test the ambient temperature inside the case. None of their tests lasted more than 10 minutes, which isn't realistic - ambient temperatures are a slow creep, and have far more of an effect over two hours+ (in the average gaming session, where the whole machine's being ragged to death). I also didn't spot any point in the video where they mention the temperature in the room.

Then you've got to contend with the fact that the average tower is not sitting on top of a desk, it's down by your feet and shoved in a corner. Add in a non-modular power supply and a much smaller case with drives cramping everything so that cables become a genuine impediment to air flow and coat them in dust, on a summer's day in a non-air-conditioned room...suddenly it really is a problem.

It's no single thing - lots of tiny little increases which you'd happily ignore in isolation. Thing is, they don't have to increase the ambient temperature much; they just have to get the internal temperature of the case up to the tipping point where the CPU and GPU coolers can no longer effectively get rid of the incoming heat. That can be as low as 32 degrees in some situations (yep, been there).

The point is that under extreme load (the hottest the CPU and GPU get) there was no change at all from filling the case with excessive cabling and only literally blocking the fans with boxes had a direct impact.

Messy cabling is inexcusable - but messy cables or a non-modular PSU wont have a direct impact on gaming performance - unless it's an extremely small case with an extreme amount of cables

digitalscream

Myranda said:

The point is that under extreme load (the hottest the CPU and GPU get) there was no change at all from filling the case with excessive cabling and only literally blocking the fans with boxes had a direct impact.

In very specific circumstances - there was no change at all, when you only test for a maximum of 10 minutes. It takes much longer than that for the effect to be measurable when you're only looking at CPU and GPU temperatures...however, when it does become noticeable then it's a runaway effect and there's limited time to act before actual damage occurs.

Again, I'll say this: they didn't measure the single most important temperature here - the ambient temperature inside the case.

Danny1969

That video is pointless, I seriously wonder why anyone would watch it. It's basic physics understanding cooling and convection. Cable management does make a difference, where you put the box makes makes a difference. Keeping the thing free of dust makes a difference

My sons gaming PC has just blown up after 4 years of hard use. The caps in the SMPS blew after the fan failed and all the 12V rails to the board went high and took that out as well. Luckily the drives and CPU survived. Don't scrimpt on the power supply, this was a £55 unit when I brought it from Novatech and I guess I got what I paid for 

digitalscream

Danny1969 said:

That video is pointless, I seriously wonder why anyone would watch it. It's basic physics understanding cooling and convection. Cable management does make a difference, where you put the box makes makes a difference. Keeping the thing free of dust makes a difference

My sons gaming PC has just blown up after 4 years of hard use. The caps in the SMPS blew after the fan failed and all the 12V rails to the board went high and took that out as well. Luckily the drives and CPU survived. Don't scrimpt on the power supply, this was a £55 unit when I brought it from Novatech and I guess I got what I paid for 

I do believe that's the first time you and I have been in agreement over computers, so I've duly awarded you a Wis

However, as to the reason people would watch it - generally, enthusiast gamers want someone to validate their approach (case cooling is often the last thing anybody thinks about or spends money on, along with power supplies, because it's not sexy and doesn't directly improve performance), so a video that confirms that it's pointless will resonate quite well and get a lot of views.

Of course, this is the Internet, where jokers with a YouTube account, good production values and a limited understanding of what's going on trump actual knowledge and experience

Axe_meister

For 1080p gaming you don't really need to spend more that £180, either a 1050ti (£150) at the budget end and will still get you 60FPS on most games, or an RX470 at £180, £30 more and you are into 1440p territory.
I even found an RX480 4GB card for £180 today. That's alot of power for the money.

ThePrettyDamned

Axe_meister said:

For 1080p gaming you don't really need to spend more that £180, either a 1050ti (£150) at the budget end and will still get you 60FPS on most games, or an RX470 at £180, £30 more and you are into 1440p territory.
I even found an RX480 4GB card for £180 today. That's alot of power for the money.

There was a GTX970 on here for sale for £120 recently. Monster card for that money. 

littlethoughts

So what I've learned is that I should make it myself, buy a £600 graphics card and fill the case with t-shirts. Thanks everyone.

Flink_Poyd

Ive been thinking of buying a gaming PC as my son plays games on the Mac more than console. I could quite easily build one myself as I had a little sideline fixing computers and gadgets. Problem is that I havent kept up on any of the specs/requirements for PC's for years, the parts picker way of doing it is just bewildering even to me and Im pretty clued up tech wise.
 What sort of specs would I be looking at for the following use?

Me. Usual browsing, some graphics/logo design along with pretending to be a rock god.

Nipper. Steam games, Minecraft/Roblox but able to run more graphics intense games as and when needed and a view to upgrade when necessary.

Off the shelf or self build doesnt matter really

 If I spec a PC myself I end up having to sell a kidney.....

Axe_meister

Min spec for 1080p gaming i3-6300 8gb Ram.
Gtx 1050ti GPU/Radeon RX 470

Next step up 1440p. I5-6500 8GB
GTX 1060/ Radeon RX480

Flink_Poyd

Axe_meister said:

Min spec for 1080p gaming i3-6300 8gb Ram.
Gtx 1050ti GPU/Radeon RX 470

Next step up 1440p. I5-6500 8GB
GTX 1060/ Radeon RX480

Thanks. I can at least start looking. Rough guess would be £550 pre built, just have to work out the self build price and convince my self to switch back to Windows.