Darren Yates [Darren] has done a good job at making a fair comparison by using UserBench suite of benchmarking software, however I have found several incosistancies and isses in the article that read strange to me and don't make sense when doing a benchmark review.
"Our Test System"
There is nothing wrong with their test system as they have built it with quality parts. However, my concern is mainly with the CPU configuration. Nothing wrong with the Intel Core 2 Duo processor, but PCUser decided to overclock it for their benchmark. Why?
It's defined by Intel themselves that overclocking is a bad idea (http://bit.ly/inteloc), resulting degration of general system reliability as well as possibility of killing the processor despite speeding up the processor's clock time.
In my opinion, these results are already tainted due to the procesor being modified beyond it's factory standards, comprimising the integrity of the processor and the tests. My angle has always been from the everyday user. Not every user overclocks their system, and CPUs don't come overclocked. So, there will be alot of people questioning wht can't mach the times published right out of the box.
I say this was a bad move by PCUser.
Starter vs. Ultimate Comparison: What the hell?
To be released with this version of Windows, consumers will see on the shelves three distinct and familar editions; Home [Premium], Professional & Ultimate. In the background we of course have the N edition (No Media Player), the almost E edition (No Internet Explorer) and the Starter edition.
The Starter edition in Windows Vista was intially marketed to emerging countries and low-end devices. Australia never really saw Windows Vista Starter Edition at all. In Windows 7, starter edition has been developed with the targets of again emerging countries and low-end devices especially the now popular Netbooks.
Currently in Australia, netbooks are packages with either propietry Linux-based operating systems or Windows XP Home. But once Windows 7 is officially released to the public, we will see it prepackaged on all the Netbook devices. So as I said, Windows 7 starter is designed for low-end, barebones machines. Not mainstream desktop or laptop solutions. But PCUser had other ideas.
In this article, PCUser decided that the comparison should be between the absolute basic edition to the absolute highest edition, Ultimate edition. Their main question was the following:
...whether ther are other limitations in Starter that may cause it to run slower than the more expensive versions of the OS.
My question is why would want to run Starter edition on a Core 2 Duo based desktop? Starter edition is (as I mentioned earlier) targeted to basic, barebones machines. But PCUser thought it test the OS on everything (multimedia, imaging, security and gaming) against the top of the line, Ultimate. I find this rediculous, a waste of time. You are not finding Starter edition on your next Pavillion. And you're not running your next LAN Gaming session from your EeePC. Their results were of no great surpirse as their test results were pretty much level proving that the different edition didnt have any hidden advantage or limitations.
My main problem is that starter edition isn't apart of the official product line, its a specialised release. So why treat it and conduct this type of benchmark on it?
"Windows 7 vs. Vista vs. XP"
I was impressed with the extent that PCUser went with this test. Here, PCUser compared the follow versions of Windows:
- Windows XP Professional SP3 (32-bit)
- Windows 7 Ultimate (32bit & 64-bit)
- Windows 7 Starter (32-bit)
- Windows Vista Ultimate (32-bit & 64-bit)
- Ubuntu 8.10 (32-bit)
- PC User UserOS Extreme 9.04 (32-bit)
- Ubuntu 9.04 (32-bit)
According to the tests, Windows XP still ranks the fastest in most tests, but then Windows 7 comes in a close second. But then my comfusion is why test 64-bit versions if you arn't going to test every 64-bit version. Windows XP has a 64-bit version, but it wasn't not mentioned. Nor was Ubuntu's. Another area where they seem to have a hole in their research. This issue was even questioned in the previous sections table of results as they listed the times of Windows 7 Ultimate 64-bit, but Starter doesn't even come in 64-bit. So how is it relelvent.
Although Windows XP 64-bit edition was ahead of it's time, PCUser should have still tested it while they were conducting irrelevent testing. This article was supposed to be informing users of Windows 7 in relevent ways, not in unrelatable methods.
"Built-in GPU Drivers"
It is widely understood that developers will through in code just to get the product out the door. Microsoft as a developer is no different. With Windows Vista, marketing and communication was rushed with it's partners resulting in very poor overall support. Windows 7 is essentially the redefined Vista. But with this in mind, I can't see Microsoft having the time and support to develop hardward drivers. Especially for graphics. In this section, PCUser test the built-in GPU drivers that are packaged with Windows 7 in comparision to those developed by the chip developers themselves. But before they even start, who would you think to have the better driver?
The question PCUser raised here was whether or not the built-in drivers compared for true 3D performance. In the same paragraph we find the answer; No they're not. What a freakin surprise. They reveal that Windows was good enough to detect and install the drivers for their test graphics cards (Geforce 8600 & Radeon HD 4850) to at least start with the correct resolution. However when it came to high-end testing, downloading the latest official drivers was the way to go. What a surprise? Don't people do this anyway?
"Is Windows Or Linux Faster?"
This was my favourite section of this entire article. Previously I had was a Linux fanboy and defended to the core of the OpenSource philosophy. But since I have since come to the relisation although Linux is very developed and does have it's markets in the world, it will never beat Windows or over take the desktop market. Never. But I always enjoy a good Windows/Linux fight. And this sections starts this fight. The test to see which one is faster.
What I was pleased about this test was that the software they are using to benchmark the system is the same on both OSs. I like consistency and this was a very good start. But this was my favourite line:
"According to our tests, the overall performance difference is about 8.5% in favor or Windows."
Now if that doesn't start shit I don't know what will. However, to fair, PCUser fails to state what distro of Linux they are using (although we do assume they are using their own PC User UserOS or Ubuntu, but it would have been niced to be disclosed). In basic processing it seems that Windows and Linux are equals but when it comes to Multimedia, Windows is out front. My other favourite line:
"It's not that Windows is more repsonsive than Linux...".
Clearly they are hurt by this (LOL).
I don't mean to be Windows bias, but the fact they even made the comparison is rediculous as Linux is so absolutley varied to Windows on all grounds that it makes it near impossible to make a fair comparison on performance.
Sorry Linux folks :S
Simply on this section, you would expect PCUser to make more of a statement than
"...that Windows 7 is a positive refinement of Vista".
I mean, come on. That's it? You mentioned about some features that should have mentioned earlier, but really, nothing actually more concrete than "It's sorta ok".
Agreed, Windows 7 is still yet to hit the masses but if the tech community has accepted so highly already than I would expect the public will be adopting Windows 7 more than they did with Windows Vista. But the problem here is this line:
"Is it as good as XP, though? If you're still happy using XP and think it Microsoft's finest hour, you may not think the same of Windows 7. This is especially true when you consider that XP ships as a 550MB CD, while the 32-bit versions of Windows 7 DVDs start at 2.4GB and the 64-bit version 3.2GB"
Their conclusion further goes on to say that although it's better than Windows Vista, they wouldn't spend $299 to upgrade to Windows 7 Home. Even though it's better out of the box then any previous version of Windows, they will stick with XP. And by the way, what difference does the size of the OS on the disc is? You install it and that's it, right? Or is the folks at PCUser mag doing something else that were missing out on?
MY final verdict is that although Darren has worked very hard to conducted thorough tests on various different configurations (and I commend him on his efforts), the article seems to miss key points while examining other irrelevent comparisons, and then introducing new points in the conclusion. My opinion of Windows 7 is that is a big imporvment on previous versions of Windows, especially since it's the first in it's line able to run on lower hardware specs than it's predecessor. Impressive. Well done Microsoft. You will win back alot of computer users with this release.