Windows 7 vs Windows 8 – Integrated Apps
Windows comes with a basic assortment of apps and software to get user started with things like entertainment, productivity and creativity. While this software isn’t going to be as robust as full featured apps, they can still be useful.
To test Windows 8 and Windows 7 app performance, we are going to use the trace-based benchmark utility PCMark 7. You might have heard about it – we have been using their software for years. Generally we only need to use the ‘overall’ scores which are calculations based on the actual test results.
This time, we are going to look at the test results themselves. Because a score of “4800″ vs “4500″ doesn’t tell you a whole lot.
First up, we’ll see how the different versions of the Internet Explorer engine handle browsing a lot of tabs at once:
This test runs three tabs, and determines how fast pages can be loaded in all three concurrently. As you can see, Windows 8 and the new IE engine will load pages faster than in Windows 7. Most people wouldn’t notice this difference in reality, but the performance is there if you ever need it.
Next we’ll use Windows’ integrated cryptography API:
In this test, Windows CNG is used to decrypt and checksum data. This time it looks like Windows 7 is slightly faster, although once again the difference is minor.
The next test uses Windows Imaging Component to perform basic operations on pictures, such as color correction, stretch, flip, and rotate:
Windows 8 receives a huge performance gain, especially on Intel CPUs.
Finally, we will use Windows Media Foundation to transcode some videos. In the first test, we transcode a 1080p 24 fps H.264 video from 12 Mbps to 10 Mbps:
This is a huge gain, and once again my theory is that this test has suddenly become multithreaded.
However, if we take the same video and add downscaling to the mix:
In a similar test, the same source video is transcoded and downscaled to a 320×240, 30 fps, 500 Kbps file, and this time the Windows 8 version performs well below Windows 7. Unfortunately I have no theories as to why this happens. Two similar tests yielding completely different results is not something we’re used to seeing.
Windows 8 vs Windows 7 Boot Time
Before moving onto desktop applications, we’ll take a quick look at boot time. For this test, we recorded the boot time starting from AFTER the AHCI initialization screen, up until the cursor becomes responsive. This test was done after the full suite of office applications were installed, to replicate a fully loaded OS rather than a brand new one.
Windows 7 already shaved a lot of time off the previous version’s boot time by giving you access to the desktop while it continued to boot in the background. It looks like Windows 8 takes that a step further while shaving an extra few seconds off. One thing you notice about booting into Windows 8 is that you are able to do things pretty much immediately, especially if you’re using an SSD.
On the next page, we’ll take a look at office application performance using another trace benchmark, SYSMark 2012.