Downloading & Trying Windows 7 Beta Public

[Solver]

That ZDNet benchmark sucked. They didn't try 64 bit systems and they ran all tests on a powerful (quad core) computer. I'd definitely also like to know how the systems compare on dual-core computers, which are the norm now, how the 64 bit versions work. Single core tests are also interesting because of netbooks.

[Asher]

They did a newer benchmark last week (no raw numbers, just relative performance).

It includes both a high end (quad core, 4GB RAM) and low end (Pentium 2.2GHz dual core, 1GB RAM).

http://blogs.zdnet.com/hardware/?p=3236

In general, Windows 7 is faster than Vista is faster than XP

[Q Classic]

Except for file copies. What happened with Vista?

[Solver]

I've seen that benchmark, it's so useless that it's hard to express. They just rank XP, Vista and Win7 as 1st, 2nd or 3rd for each test. I need numbers, a benchmark needs numbers. 7 moves small files faster than XP, which moves faster than Vista. So, by how much? Is that by 1%? 100%? Maybe there's a test on which times for all three systems were within 1%, making the first-second-third gradation pointless? A performance comparison with no score numbers / times is truly something...

[Asher]

I've seen that benchmark, it's so useless that it's hard to express. They just rank XP, Vista and Win7 as 1st, 2nd or 3rd for each test. I need numbers, a benchmark needs numbers. 7 moves small files faster than XP, which moves faster than Vista. So, by how much? Is that by 1%? 100%? Maybe there's a test on which times for all three systems were within 1%, making the first-second-third gradation pointless? A performance comparison with no score numbers / times is truly something...

Why does it matter how much faster, if it is faster?

I'm not sure why you're spending so much time *****ing so loudly. The frickin OS is free to download and try. If you want to see how fast it is, then do it?

[snoopy369]

Do they generally turn Aero (etc.) off on benchmarks? I would have to wonder if one of the problems on Vista is that, once you run the performance checks, it by default activates Aero (etc.), which kills performance. I was quite surprised to find Aero activated just because I ran the performance checks (I don't recall being asked to turn it on, anyway; it must have been active pending performance checks being run, or something similar), and my performance suffered considerably (I'm on what probably qualifies as a mid-range system with a mid-high graphics card). I of course turned it off, and regained my performance, but I wonder how many people have problems because of it and don't know it's the cause of the problems...

[Asher]

Do they generally turn Aero (etc.) off on benchmarks? I would have to wonder if one of the problems on Vista is that, once you run the performance checks, it by default activates Aero (etc.), which kills performance. I was quite surprised to find Aero activated just because I ran the performance checks (I don't recall being asked to turn it on, anyway; it must have been active pending performance checks being run, or something similar), and my performance suffered considerably (I'm on what probably qualifies as a mid-range system with a mid-high graphics card). I of course turned it off, and regained my performance, but I wonder how many people have problems because of it and don't know it's the cause of the problems...

I've never, ever, seen Aero hurt performance on a properly configured "mid-ranged" or higher card. And we've done tons of testing here...

It's the difference in software rendering versus hardware rendering. Are your drivers old for your video card? Disabling aero will render the windows and your desktop on your host CPU with your system RAM, enabling aero will render the windows on your idle GPU with your idle GPU RAM, freeing up CPU resources...

Now if you have a misconfigured driver or an old one, I can see it might cause problems...

[Solver]

Why does it matter how much faster, if it is faster?

I'm not sure why you're spending so much time *****ing so loudly. The frickin OS is free to download and try. If you want to see how fast it is, then do it?

Because I would care about the practical benefits of an upgrade? If a new OS can do stuff 1% faster, that's nice to know but I'm not going to be all excited. If it can do stuff 20% faster, or more, I'll be eager to upgrade. If Windows 7 is really noticeably faster than XP and Vista, then MS really have a hit here. And my beef is with ZDnet's reporting in this case, not MS if it isn't clear.

But I guess I am going to try Win7 soon, at least on a virtual machine to check the new features, even if a VM won't give me a fair estimate of performance.

[Asher]

Honestly, it's not significantly faster than Vista. The people who think it is have a whole lot going on to talk about (Mojave effect, fresh install, comparing Vista pre-SP1 to Windows 7, placebo, etc).

It's why there's not a whole lot of benchmarks for it, they're boring.

[Solver]

Well, that's why I wouldn't mind seeing some more reliable and detailed enough benchmarks. To see if Win7 offers merely a negligible performance improvement (coupled with placebo effect) or something more significant. And I'm still curious about single-core performance. I'll be getting a netbook soon, probably won't have Windows on it at all, but I wonder if Win7 would be at all viable on it.

Also, why do MS take so long to add the really good features they already have to OSes? PowerShell is a great feature not included into Vista by default. MS has had virtual desktops, which are very popular with Mac and Linux users, but they're still a powertoys-only feature. Vista still uses the pathetic built-in calculator, whereas MS Power Calculator is actually a good tool.

[Asher]

Windows 7 actually has the new calculator.

PowerShell was completed after Vista ships and is being included in Windows 7 as an optional component (not included by default, but ships on the disc).

MS has UI issues with virtual desktops that would confused many users, so they just provide it as a free download instead.

[Solver]

I think they could've included PowerShell in Vista SP1 and XP SP3. It's quality and was finished by then. And I'd guess users could handle virtual desktops... even if not, the problem is that too few people know about PowerToys and the like. It's not something the average user will get, yet I think the average user would enjoy extra usability from virtual desktops.

[Nikolai]

Installed it earlier today. Looking good so far. Works like a charm, next to no bugs seen so far. Only bugs I've encountered, are that I wasn't allowed to make a smaller partition than ~300 GB, and that I'm unable to remove the read-only option on folders.

[CrONoS]

On a side note, people & myself are experiencing huge increase of FPS in Direct3D games on Intel Integrated. Since the GMA 965 suck as a gaming card, having an increase of 30-100% of FPS is appreciated

[Asher]

That is 100% attributed to drivers, not the OS.

The DirectX 9 and 10 runtimes in Windows 7 are 100% identical to that in Vista.