by Michael S. Kaplan, published on 2009/07/24 10:01 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2009/07/24/9847607.aspx
You may have heard that Windows 7 RTM'd the other day....
There were two machines of mine that I hadn't yet tried to upgrade as I was waiting for OEM support first.
One of them was my Dell Precision Mobile Workstation, the one I mentioned in Pretty damn close to top of the line.
But I had a spare drive so in the spirit of RTM I decided to try installing and clean and seeing what happened.
Now although the 32bit install was pretty smooth when I put Vista on it, the 64bit install was tough since I had to find drivers myself (Dell wasn't supporting it yet so they had none on their site), asnd I figured it would be a good comparison to see if the story was any better this version.
You want to know what? It was! All but six drivers were right in the win7 image ("inbox" drivers), five of the remaining six six were detected and found on Windows Online, and the sixth was just the Touchpad (the Vista driver worked perfectly). I was very quickly up and running!
Even the modem was there, which last time wasn't available at all until long long after....
The funny thing was the Windows Experience Rating!
You may recall that the Vista one on this machine was 4.7:
with a breakdown of:
| Component | Subscore |
| Processor | 5.2 |
| Memory (RAm) | 4.7 |
| Graphics | 5.9 |
| Gaming graphics | 5.8 |
| Primary hard disk | 5.0 |
Well, the Windows 7 numbers were quite a shock!
It was 5.1 for the Windows Experience Index:
With the following breakdown:
| Component | Subscore |
| Processor | 5.8 |
| Memory (RAm) | 5.8 |
| Graphics | 5.9 |
| Gaming graphics | 5.8 |
| Primary hard disk | 5.1 |
Wow -- identical hardware, 31 months later, and the scores are so much better!
I went ahead and ordered another hard drive from Dell (both the original and the replacement were 5400 RPM drives so obviously that score would be pretty easy to raise!), but the improvements -- to the processor and the memory -- are kind of confusing to me.
It was not running with or without Aero (in both cases the numbers were identical with and without Aero).
Was it a flaw in the original tests?
Was the slowdown due to Vista itself? Or the drivers on it?
Or were the new drivers simply faster in some unanticipated way so that even on identical hardware the difference was visible?
Or did Microsoft simply alter the nature of the tests themselves? It isn't like they guaranteed them to be identical tests across different versions.
I'll let you know how things go after I get the drive, but does anyone have an opinion about the improvements to the two scores?
Wesner Moise on 24 Jul 2009 11:10 AM:
The big changes were in memory and processor, which makes sense because in the 64bit version of Windows, you are using essentially a different processor and can access an additional megabyte of RAM.
1) You are using the native 64bit mode of the processor, which performs better than the 32bit mode.
2) You can only utilize about 3GB of your 4GB of installed RAM in 32bit mode; the rest of the address space is taken by video memory, etc. But in 64bit mode, the address space suddenly becomes "unlimited" and your machine can address all available installed RAM.
As for the small blip in hard drive performance, that could easily be due to additional Windows 7 optimizations.
Michael S. Kaplan on 24 Jul 2009 11:36 AM:
Both are actually running under 64bit (the Vista run was after I put the drivers together), note the system type.... :-)
Russ on 24 Jul 2009 2:12 PM:
You guys changed the tests for Win7
http://blogs.msdn.com/e7/archive/2009/01/19/engineering-the-windows-7-windows-experience-index.aspx
htd on 24 Jul 2009 3:53 PM:
honestly, I dont think that test is that useful and I dont think many people really care about it.
Michael S. Kaplan on 24 Jul 2009 4:44 PM:
That article is interesting, though doesn't really explain higher scores (all of the score change examples were about lower scores).
Tom Gewecke on 24 Jul 2009 5:34 PM:
I'd be curious how it runs on your Mac (if that is possible).
From what I've seen written so far the internationalization improvements seem substantial.
http://m10lmac.blogspot.com/2009/07/new-language-features-in-windows-7.html
Michael S. Kaplan on 24 Jul 2009 6:21 PM:
Regular readers will recall that for Vista the MackBook Pro had a 0.1 lead on the Dell when it came to the Windows Experience Index; I admit I am curious too....
I'll be upgrading the MacBook Pro as soon as I make sure there are no blocking driver issues to worry about (I did the same effort to support 64bit Vista on the Mac -- roughly six months before Apple released an easy to use driver pack -- and it was a huge pain in the butt!).
Jeroen Vos on 27 Jul 2009 2:37 AM:
I think its because the highest value you could get on Vista is 5.9, and the highest you can get on Windows 7 is 7.9. I guess they had to scale things a bit.
Probably memory hasn't had that much of a performance increase in the last months, but graphics and processors have. To keep memory from always being the slowest. They probably had to tweak the ratings here.
GordonSchumacher on 27 Jul 2009 5:26 PM:
"I think its because the highest value you could get on Vista is 5.9, and the highest you can get on Windows 7 is 7.9."
I'm waiting until it goes to eleven!
Yuhong Bao on 28 Jul 2009 1:14 AM:
"Both are actually running under 64bit (the Vista run was after I put the drivers together), note the system type.... :-)"
Yep, to use more memory than the amount that can be mapped in 32-bit physical address space:
1. the OS have to be able to address more than 32-bit physical address space via 64-bit or 32-bit + PAE
2. the memory controller must also be able to actually map memory over 4 GB (the memory controller built into the 945 and older northbridges can't)
3. the BIOS also must configure the memory controller to actually do so, and list the additional memory mapped over 4 GB in the e820 address map.
Igor Levicki on 2 Aug 2009 12:12 PM:
Btw, my scores with RC on my workstation machine are as follows:
CPU: 7.4
RAM: 7.7
AERO: 6.0
Game: 6.0
HDD: 5.1
GPU is an old 7950GX2 which will soon be replaced.
McDowell on 4 Aug 2009 7:13 AM:
Apologies if this is the wrong place, or this has come up before, but does 64bit Windows sound the death knell for OEM code pages on the console? I haven't run a 64-bit version of Windows, so I can't check if I'm just behind the times.
I can see the point of cmd.exe defaulting to a DOS codepage for DOS compatibility, but if the operating system no longer supports 16bit apps, it might be a good time to change. I live in hope that, even if Unicode isn't on the table, it might one day default to the system default "ANSI" encoding. (Yep, I'm aware of TT fonts/chcp and PowerShell.)
Or are there deeper issues of which I am not aware? I've never had cause to go near the recovery console and I know the rules change on full-screen.
Michael S. Kaplan on 5 Aug 2009 10:28 AM:
Hey McDowell,
Very very offtopic :-) but I moved the conversation over here....