There's a lot to like about Windows 7, not least its many improvements over Vista: the new OS is faster, less demanding on resources, has better designed security and contains many new productivity-boosting features.
If you were an early Windows 7 adopter, though, you may already have noticed that one old problem still remains. The more you use your PC, adding and removing applications, the more junk builds up throughout your system, and the slower and more unstable it eventually becomes.
You need to treat the problem, detoxing your PC on a regular basis to remove the leftovers – but how, exactly? Which areas of Windows 7 are most susceptible to this gradual degradation? Are there any tools or benchmarks you can use to reveal problem areas? How much can all this clutter slow you down, anyway, and what's the best way to remove it all and restore your system to its optimum performance?
As we researched this article, one point was clear. Windows 7 is very different internally to Windows XP, and we couldn't simply assume that old tricks, like optimising services, would work in the same way. What we needed to do was design a test, something that would reveal exactly why Windows 7 systems slowed down over time, and help uncover the best way to restore that initial new PC performance. And so that's exactly what we did.
Designing the test
We started our trial by obtaining a powerful new 3XS Intel X58 Core i7 PC from Scan Computers. The machine featured a quad-core Intel Core i7 920 (which was overclocked by 20 per cent), 6GB of RAM and a speedy SATA 300 Samsung hard drive. It was an excellent performer that we knew wouldn't choke unless it was faced with a set of major performance problems.
When the 3XS PC arrived, we installed the latest Windows 7 (Ultimate Edition, 32-bit) and driver updates and then set about establishing baseline measurements of our PC's performance. The best Windows boot time – which we're defining as the time that elapses between the 'Starting Windows' message and the desktop appearing – was 22 seconds.
Seeing the desktop means nothing if you can't use it, so we also measured the time between the 'Starting Windows' message appearing and the point that we were able to launch IE and have it display our Google homepage (28 seconds). We also used Task Manager to collect data on free memory and system activity (processes, threads, and so on).
Finally we checked how long it took to launch apps, including Firefox and Outlook (both around four seconds). With the performance of our clean system safely defined, we set about abusing it.
We installed Windows Live tools, iTunes, Adobe Reader, browsers, antivirus apps, Microsoft Office, DVD-burning suites, video-editing tools, a large Outlook inbox, hundreds of fonts and more. We accepted every extra that was on offer, then reinstalled and updated the apps before moving plenty of files around to ensure hard drive fragmentation. And what did this do to the benchmarks?
The plain Windows boot time increased by around a third, from 22 to 30 seconds. Our system was unusable after that for a long time, though, with IE not displaying Google for 140 seconds. Task Manager showed that system activity had more than doubled. Outlook now took five times as long to launch (21 seconds), and shutdown time increased by 50 per cent to 18 seconds.
So even a powerhouse like our 3XS system can be seriously affected by clutter. Now our really important tests began: discovering how to reverse this slowdown.
Defrag options
The hard drive is a big bottleneck on most PCs, and defragging has traditionally been one way to boost performance. Windows 7's own defrag tool completed the task in a little over 20 minutes, confidently reporting that there was now 0 per cent fragmentation. But this had little effect on our PC, shaving one second off boot time and leaving other benchmarks unaffected.
We weren't convinced, and ran Auslogics Disk Defrag immediately afterwards. This produced some interesting information: it thought our drive was still 16 per cent fragmented. We told the program to optimise our file layout (go to 'Settings | Program Settings | Algorithms | Move system files to the beginning of the disk') and set it to work.
This delivered real benefits. Boot time fell from 29 to 26 seconds; IE was usable after 107 seconds, a 23 per cent improvement; and launch time for Outlook fell by a third.
We can't guarantee you'll see similar results, as every defrag situation is different, but it's clear that Windows 7's defrag tool alone won't necessarily do the job. We advise you click Start, type defrag, click 'Disk Defragmenter' and make sure that scheduled defrags are turned off for the moment.
Then install Auslogics Disk Defrag, turn on the option to relocate your system files, click 'Settings | Program Settings | Schedule' and set it to run every few days to keep your drive running optimally.
Your comments (12) Click to add a new comment
andyprice
May 16th 2010
12. I thought it was a good article overall and shows how installing lots of extra programs and add-ons really kills performance. It also disproves the common misconception that modern PCs are so powerful that they can handle anything you throw at them.
Alert a moderator
mike_williams
May 12th 2010
11. @psikys, we all get overheated sometimes (me included), but not everyone has the courage to admit that later. So thanks, & don't worry about it.
Re: the SSD, that was a part of the original brief to me. I had to look at various ways of decluttering your PC (which inevitably led to me covering some very well-trodden ground, as you say), then measure the improvement you'd get from adding an SSD to your system.
Should we have included ReadyBoost and other low-cost alternatives, as well? Maybe, there would definitely have been some value in that, though it would also have reduced the space we had to talk about SSDs: there's always some trade-offs and compromise in print features.
Alert a moderator
psikys
May 12th 2010
10. Okay. Came here today with the sole purpose of reading any new comments. Mike, I apologize for being kind of a jerk last night, but the 4gb(video ram included) vs 6gb thing killed me.
Anyway, I finished reading your article. I agree that all of these things will help your Windows install, but the fact remains that nothing you suggested here really changed from what we've done with Windows XP (not that it needs to, but just making that statement because of your statement about how different 7 is from XP). Cleaning your start up items, cleaning your services, defrag, file clean up, registry clean up, all very common fixes.
I feel that this article definitely has some great points, but my concern is you suggested a hardware fix but went straight for the expensive one. You did however leave out a GREAT one, ReadyBoost. A simple 15 dollar performance increase vs a 150 dollar hard drive change which many readers may not be comfortable performing themselves, therefor increasing the cost of the performance increase even higher.
Overall, now that I have my head out of my ****, good article. But not complete.
Again you have my apologies Mike!
Alert a moderator
mike_williams
May 12th 2010
9. I see there are lots of people understandably asking why I used a 32-bit OS on a system with 6GB of RAM for this piece. Let me explain.
Originally the article was commissioned for PC Plus on the topic of "detoxing your PC", regarding ways to remove unwanted clutter and other optimisations that could restore your system to its "as new" performance (or as close as we could get).
The magazine wanted plenty of practical tests and benchmarks. I'd clearly need a test PC, and as time was very short I said I'd use one I was buying for myself at the time, which was the spec as described, with 6GB of RAM, because I planned to install 64-bit Windows 7 later.
For the purposes of the feature, though, I opted for 32-bit Windows as it's what most people have, and I wanted the test results to relate to the maximum number of people, without introducing new variables like whether we were using a 32 or 64-bit app.
So, the amount of RAM is a red herring, really. Had I obtained a test system for this article alone I'd have used 4GB, and the only reason it had 6GB was for my personal use afterwards. In retrospect, that should have been explained in the piece, or simpler still, I should just have removed 2GB before I started: mea culpa.
Alert a moderator
b775854
May 12th 2010
8. I almost chocked when I saw "IE"...
Alert a moderator
psikys
May 12th 2010
7. Okay. First off, I expected a lot more from TechRadar. Let me start off with this, I am a Technical Trainer and Curriculum Designer for a major online educational institution in the US. I work with multiple systems every day spanning from Linux, Mac, Windows and numerous hand-held devices.
What is the first thing I noticed in this article? First page? 6GB of RAM on a 32 bit system. At best (as madjedi has already stated) the designers have set this system up to use roughly 55% of it's available system RAM. This is one of the most simple concepts of PC building. They build a machine that is designed to operate under amazing conditions and then put a stranglehold on it...I don't understand. I agree with you madjedi - I cannot trust this article.
I seriously have to question if the writers of the article even understand WHY installing multiple applications and making change after change to the OS and boot environment change boot times.
In short, bad form TechRadar. Bad form. Next time you run a test and post your findings online, please have the knowledge and foresight to run a PROPER test.
Alert a moderator
chillzatl
May 11th 2010
6. Defragging, unless you're using a program that will move system files to the front of the disk (built-in defrag does not do this), is a waste of time on an NTFS partition.
Most of the native services that the author stopped have little to no impact on load/start times.
The single biggest change one can make is removing all of the systray items that load on startup from 3rd party apps using Msconfig.
Also, the authors use of Win732 over 64, on a system with 6gb of ram is noobish.
Otherwise a decent article for the unwashed masses.
Alert a moderator
feedback
May 11th 2010
5. Editors: please (please) provide a single-page view for articles. Who wants to click to view 4+ pages? You're definitely losing readers this way. Thanks!
Alert a moderator
paul
May 10th 2010
4. D'oh - we've corrected that - it should be 6GB of course :)
Alert a moderator
bradavon
May 9th 2010
3. Agreed it's weird to be installing Windows 32-Bit (except if it's XP or on a Netbook) and clearly that's a typo, the author means 6GB.
Otherwise it's a well written article and I agree with most of it. Personally I stick with the Vista/Win7 Defrag tool. Sure it's not the best but it runs once a week and never bothers me. It also has no services or any other junk bothering when it's not running. Technically it has one service but I set it to Manual.
I used to use Diskkeeper but it had so much going on in the background, it actually made my hard disk more active and therefore slow the PC down.
That's good advice about setting iTunes to Automatic Delayed. Most of the other stuff can be set to Manual though, meaning it won't load at all, unless you need it.
Alert a moderator
madjedi
May 9th 2010
2. Ps just noticed that the system seems to have 6MB oF RAM not 6 GIG. This opens a whole other set of problems with the article!
Alert a moderator
madjedi
May 9th 2010
1. Ive not yet read the whole article but to me they seem to have failed at the first hurdle. Why install a 32bit os on on a machine with 6 gigs of RAM. At best it will only see about 3.5 gigs and probably a lot less than that. How can trust the rest of the article if they have got such a simple thing wrong?
Alert a moderator
Tell us what you think
You need to Log in or register to post comments