Rock Stable: How to Get Your PC in Tiptop Condition

PreviousNext

Introduction

Stable Gaming

October 31, 2011

Introduction

Stable gaming. Sounds like a dull topic, right? Not so. Stable gaming is, in reality, the Holy Grail of every PC gamer. Unfortunately some PC gamers tend to think that most games are buggy and unstable by default, and there's not much that can be done about it. This isn't true at all. From personal experience I can tell you that it's entirely possible to get virtually any game running with stability, as long as you tend to your own PC. So if you're having problems, don't fret. With a few simple tweaks, the ability to game with stability, unlike the Grail, will be within your grasp.

How Stable is Stable?

Different people seem to have different ideas of what the word "stable" means when it comes to PC gaming. Let's push away all those other ideas and get something clear in our minds: stable gaming means that you switch on your PC, you load up a game, and you play it for as long as you like without it crashing, freezing, glitching or otherwise making itself unplayable in any way. That's our goal.

Let's have more of this

Let's have more of this

...and less of this

...and less of this

Now I don't want to make it sound like it's all your fault if your games crash or have problems. Sure, some games have known bugs and quirks. But over many years of PC gaming I've yet to come across more than a handful of games which are truly so unstable that they can't be played for as long as I want without them going belly up at the first chance, especially after the first couple of patches are released. If you're having problems with smooth gaming, such as stuttering or lag, that's another thing altogether and is covered in my How to Get Rid of Lag article.

It's time we looked at some ways of getting your PC into shape so that it can last the distance with any game.

The Basics

You might know the drill, but let's run through it anyway. Here are several things you absolutely must do on any PC to make sure it's up to the task when it comes to gaming:

  • Update your drivers: Don't just install the latest graphics drivers (which for NVIDIA users can be found here), also make sure you grab the latest drivers for all of your other devices. Flaky audio drivers in particular are a known cause of gaming problems. Check the manufacturer's website for each and every one of your PC components. Even your router and DVD drive could do with some firmware updates.
  • Patch your games: In this day and age, when most games auto-update or at least let you know when a new update is available, there's really no excuse for not having your game patched up to the latest version.
  • Optimize Windows: Whichever version of Windows you're running, it's the foundation for your system. Anything wrong with the way Windows, your BIOS, or your hardware is configured and guess what - you're going to have problems. Run through this Windows Optimization Guide which comes in a handy PDF format.

I'll bet you dollars to donuts that you'll tell me you've already done these things. OK, I'll take your word for it, but I had to make sure. We'll move on to looking more closely at the most common causes of gaming instability.

Software Conflicts

All software undergoes testing before it's released. During testing the developers try to make sure that they cover their bases with all the different systems and hardware configurations on which people might run the game. Most of the time they do a good job, and yes some bugs get through, but that's inevitable given how complex games have become and the huge variety of PC hardware out there right now.

But there's something that's virtually impossible to do compatibility testing for: the other software that may be running in the background at any time along with a game. This is because there are literally millions upon millions of combinations of different software that could be running on a system at any one time - from the good to the bad. In many cases you may not even know they're there, because a lot of software will try to run bits and pieces of itself as part of your Windows startup process, quietly loading up things in the background which you don't really need. The formal name for a certain combination of software which causes problems is a Software Conflict, and it's much more common than you think.

The first thing you must do is install and run a good malware scanner like the free Microsoft Security Essentials to see if you have any hidden malicious software. You might also use the free Spybot Search & Destroy for another scan just to be sure. Once you're sure your system looks clear of malicious software, here's the important part: you need to either temporarily disable your malware scanner(s) whenever you play a game, or better yet, configure them for minimal background intrusion as covered in the PC Security chapter of the Windows Optimization Guide I linked to earlier. This is because malware scanners running in the background are a known cause of software conflicts, especially during the installation of games, and also while playing them.

Next up, in Windows Vista or Windows 7 click the Start button and type "MSConfig" (without quotes) in the Search box and press Enter; in XP, do the same thing in the Run box. In the System Configuration tool which opens, go to the Startup tab and you'll see a list of startup items. Click the Services tab and tick the 'Hide all Microsoft services' box, and once again it's the same story. Some of these must go, but how do we work out which ones to axe?

A nice short list of startup items

A nice short list of startup items

Expand the Command column in the Startup tab of MSConfig, and look at the filename at the end of the directory path for each item. Then enter that filename into Google, or search for it in this Startup List, and you should be able to tell both which program it relates to, and what it does, hence whether it's really needed or not.

Once you've figured out which startup item relates to what program, to safely disable it, first launch that program and see if it has any options such as 'Load at startup' or 'Load with Windows' and untick this. For example, the Steam gaming platform has a 'Run Steam when my computer starts' option under Steam Menu>Settings>Interface. You don't really need to have Steam starting up with your computer and always running in the background, you can launch Steam whenever you want to use it. Some programs however absolutely need their startup items to launch with Windows. The Microsoft Security Essentials program for example needs its msseces.exe file to load at startup or it won't launch properly afterwards.

If you can't find an option within the program itself to disable its startup item, then you can do so temporarily by unticking the box for that item in the Startup tab of MSConfig. The next time you reboot Windows, the items you unticked will be disabled and you can test to see if this affects the functionality of those programs or has any other odd side-effects. To undo these changes, open MSConfig and under the General tab tick the 'Normal Startup' box, click Apply and reboot. You can run through a similar procedure for finding and disabling unnecessary Services: look under the Services tab in MSConfig, tick the 'Hide all Microsoft services' box, then search for each service name in Google to see what it does. Untick it to temporarily disable it, and reboot and to see the effects in Windows.

Let's have more of this

A riskier method, but much more thorough

For more advanced users, you can find and temporarily disable or permanently remove a much wider range of startup items beyond those seen in MSConfig, such as unnecessary drivers and shell extensions, by using the free Autoruns utility. Be aware though that you can cause yourself some serious problems if you disable the wrong things in Autoruns, so follow the instructions under the Startup Programs and Services chapters in the Windows Optimization Guide I linked to earlier.

If you're still not sure about the value of removing startup programs and services to gaming stability, don't take my word for it - see this Microsoft Article which says the same thing. It really does make a difference. And it's worth doing it regularly I might add, because almost every new piece of software you install will probably add a startup item and a service or two to your Windows startup. That's just the way they roll these days. Even if they're not the source of your gaming stability woes, stripping out unnecessary background programs will speed up your Windows startup time, and free up memory for more useful things.

High Temperatures

Without even a hint of a doubt, I can say that a reasonable proportion of problems stem from overheating. A lot of this is due to overclocking, but even if you don't overclock, your hardware components can still overheat. They need plenty of cool air to operate within a safe and stable temperature range, and things can get hot in a hurry inside that stuffy PC case, especially during gaming. You see, people often forget the fact that gaming is one of the most strenuous things you can do on a PC. While you're having fun, your PC is doing it tough. So let's spare a thought for the suffering PCs of this world and help make their life a little easier.

The first and most important step involves checking to see exactly what temperatures your main components are reaching during your gaming sessions. By "main components", I'm referring to your CPU and your GPU (graphics card), since these two between them do the bulk of the work during gaming, and also put out the most heat in any PC case. In other words if either of these guys are overheating, the effects spill over onto other nearby components which start to get hot as well.

You'll need to download these handy free tools:

  • The GPU-Z utility is for graphics cards.
  • The Real Temp utility is for Intel CPUs.
  • The Core Temp utility is for Intel and AMD CPUs.

Install and launch GPU-Z, and then go to the Sensors tab of the utility. Here, in the first 'GPU Temperature' box you can see the current temperature of your graphics card, and it should be relatively low since you're only using the Windows Desktop, which isn't particularly strenuous - hence the 'GPU Load' box at the bottom might show only a few percent at most. Make sure to tick the 'Continue refreshing this screen while GPU-Z is in the background' box, then minimize GPU-Z into your System Tray/Notification Area.

GPU temps are nice and low, but we're only on the Desktop

GPU temps are nice and low, but we're only on the Desktop

Now run your most strenuous game, or the one which is the least stable, and play for at least 10 minutes; the longer the better. Quit the game when you're ready, and maximize GPU-Z. Looking under the Sensors tab, click the small black arrow in the 'GPU Temperature' box and select 'Show Highest Reading'. This will give you the highest temperature reached on your graphics card during your gaming session, denoted with a little green 'Max' next to the number. You can also view the lowest and average temperatures. Write these all down and close GPU-Z.

CPU temps vary a bit between the cores - this is normal

CPU temps vary a bit between the cores - this is normal

Next up, install and launch Real Temp or Core Temp as relevant. On the main page you can see the temperatures for each core of your CPU. Once again, minimize the utility, then launch your most strenuous or least stable game and play for a reasonable period. Quit, come back and check the readings under the Max. column in Core Temp, or the Maximum section of Real Temp - these show the highest temperatures reached on each of your CPU cores while the utility was open, and on Real Temp, the exact time this occurred. Note down all the minimum and maximum temps.

Here comes the hard part: working out whether the temperatures are normal or not. The reason this is difficult is because each type of hardware uses a particular architecture which may be designed to run hotter or cooler. As a very general rule, if your CPU or GPU is getting close to or exceeding 100 degrees Celsius (212 degrees Fahrenheit), and you can hear the fan working very loudly under load, then it may be overheating. Most modern GPUs and CPUs will automatically throttle down their performance when they get past their safe temperature range to prevent permanent damage. However continually running a CPU or GPU at very high temperatures can lead to reduced performance, instability, and will shorten the lifespan of your component as well. The only way to really know what's safe and what's not is to read various reviews for your component to see what temperatures were recorded at idle and load under test conditions, and to also Google the particular model name of your component along with the words "safe temperatures" (without quotes) to see other peoples' results. According to NVIDIA, for a high-end GPU (eg. GeForce GTX 580), operating temperatures between 80-90 C are fairly normal.

One other factor to consider is that even if your temperatures are hovering below the danger mark now, what will happen on a particularly hot day, or when a new and more strenuous game comes along? Always leave a fair bit of headroom when it comes to temperatures.

If you are running higher temperatures, it's time to put some heat-reducing measures in place:

  • Beware the Overclocks: I consider this the Golden Rule - If you're overclocking, always start by assuming your overclock is the source of any instability. Overclocking is undoubtedly the single largest cause of issues on most PCs, because it's easy to do, but extremely hard to do correctly. Reduce your overclock to reduce your heat and increase your stability.
  • Banish the Dust: Because a PC case has several fans sucking air into it, it can quickly gather a ridiculous amount of dust, and this blocks the flow of air into and out of the case. If you haven't opened your case before, or at least not recently, be prepared to see a whole lot of dust bunnies. Shut your PC down, turn off the power, then gently clean everything inside with a barely damp lint-free cloth, q-tips and even a can of compressed air if available to you. This step alone has been known to resolve many a mysterious case of instability.
  • Cable Clean-up: One relatively easy procedure to improve the airflow in your case is to tidy up your cable configuration. Make sure no cables are blocking air paths to your CPU or GPU. Group cables together as much as possible and then clip or twistie-tie them out of the way to allow the air to flow more freely around your components.
  • Upgrade Time: If you have a cramped and poorly ventilated case, instead of spending big bucks on larger fans and aftermarket cooling, a smarter move is to buy a bigger case with more airflow potential and better fan placement.
  • Improved Cooling: The majority of problems will be resolved by the measures above. But if nothing else works, advanced users should consider buying a better fan and heatsink combination for the CPU or GPU. If you already purchased one but it doesn't look like it's doing the job, then take it off and reseat it again. When re-applying thermal compound, don't improvise—follow the instructions provided; too much or too little thermal compound can cause overheating.

If you're unfamiliar with how to safely handle computer hardware, and if you need more cooling tipsm see the Hardware Management section of the Windows Optimization Guide.

Keeping temperatures down requires constant attention to all of the factors above. Even advanced users can forget to regularly clean out the dust build-up in their case; or neglect the fact that their overclocks, which may have seemed stable up till now, will trigger instability in the latest game they've purchased; or they may just plain mess up the installation of their new aftermarket GPU cooler and not notice the temperatures skyrocketing during a gaming session. Get into the habit of checking temperatures regularly.

Hardware Instability

Aside from heat, there are times when hardware just isn't up to the task. By that I mean that a component may be on its last legs, physically faulty and unable to perform properly. When an electronic component has one foot in the grave, it usually starts producing errors, and instability results. To see if your hardware may actually be physically faulty, first remove any overclocking on your components, then look for the following symptoms:

The Windows BSOD can be the hardware equivalent of the Grim Reaper

The Windows BSOD can be the hardware equivalent of the Grim Reaper

GPU: Aside from crashing and freezing, you'll frequently see strange glitches such as flashing or missing textures, sparkling dots, and strange colors in games. If this is occurring at normal temperatures and other people aren't reporting the same problems in those games, it's likely your GPU has had it. A GPU stress test like the free Unigine Heaven will really bring out these problems.

CPU: A dying CPU will result in all sorts of data errors during normal Windows usage, as well as the entire system often freezing. If temperatures are normal, then a CPU stress test like the free Prime95 will typically fail within the first few minutes on a faulty CPU.

Memory: Similar to a dying CPU, there will be data errors within Windows, as well as sudden reboots if your RAM is faulty. Games will also suddenly crash to desktop with no warning. Windows Vista and Windows 7 have a built-in memory testing tool which you can access by typing "Memory Diagnostic" (without quotes) in the Start>Search box; on Windows XP use the free Memtest86+. These tests run outside the Windows environment, making sure that they only check the RAM for physical faults, not incorrect Windows settings. If your memory isn't overclocked in any way and fails the test, one or more sticks are likely faulty. Rerun the test one stick at a time to find which one it is.

Hard Drive: Often forgotten is the humble but important hard drive. Use a utility like HD Tune and check under the Health tab to see if there are any detected errors. Also run the Windows Check Disk utility by right-clicking on your drive name in explorer, selecting Properties>Tools then clicking the 'Check Now' button and clicking Start, then rebooting to allow it to run. Numerous bad sectors will indicate an unreliable drive.

Power Supply: This is both a critical area for stability and yet also a tricky one to test. There's no easy way to know if your Power Supply Unit (PSU) is faulty, or just not up to the task. You should first make absolutely sure your PSU can supply enough watts for your setup by using a PSU Calculator. Then take a look at this PSU FAQ to get an idea of the other factors you should take into account; it may be time to invest in a better PSU. In particular if you're running a multi-GPU setup, check to see if the PSU is SLI Certified. NVIDIA maintains a list of certified SLI power supplies here.

These tests are not an infallible way of detecting a faulty component, but they certainly set you on the right path to finding out how sick or healthy your PC actually is. From what I've seen, the average PC component may have an error-free life of perhaps 5-7 years at most before something inside it malfunctions and it becomes useless. They're not really designed to last.

Rather than keeping them warm and making their last moments as comfortable as possible, as soon as you diagnose a dying component you need to ditch it ASAP. It's usually not cost effective to repair a faulty component. If it's under warranty get it replaced, otherwise fork out the dough for a new one fast, else you will see all your data eventually corrupted, and it may even creep its way into your backups - things will get very frustrating very quickly.

Clean Reboot and Conclusions

Clean Reboot

One quick tip for those who use one of the Windows Sleep modes when shutting down their PC, or who don't shut down their computer at all between sessions: do a clean reboot often. Unlike a normal shutdown/restart process, the Sleep modes, as well as PCs which don't get shutdown at all, start to accumulate a lot of detritus in memory. Windows memory management usually handles things quite well, swapping out and eventually removing the portions of programs it doesn't need. But it isn't a perfect process, and just as having lots of background programs can lead to software conflicts, having lots of leftover resources from various programs increases the potential for instability. All you need to do is restart your PC every few days, or after any crash or odd behavior, to reset the memory state afresh.

Conclusion

If you're having problems with a game, don't automatically lay the blame for any instability on the game itself, your drivers, Windows, or what phase the Moon is in right now. I've been gaming for over 20 years, and let me tell you brother, games aren't any buggier than they used to be, despite what you may hear. Getting things stable on a PC is not a magic trick, it just takes some basic checking and maintenance. It may not be a lot of fun, which is probably why a lot of people forget to do it, but it's necessary. But once you have the right knowledge and stick to a few good principles, and you'll be on your way to PC gaming nirvana.

submit to reddit Email