Is Windows 7 Faster Or Just Smarter? 619
Barence writes "The Windows 7 unveiling garnered largely positive coverage, with many hands-on testers praising it for being faster than Vista. But is it actually? To find out, this blogger ran a suite of benchmarks to see just how much quicker Windows 7 really is — and the results weren't quite what he expected. 'The actual performance gap between Vista and Windows 7 is ... nada. Absolutely nothing. Our Office benchmarks and video encoding tests complete in precisely the same time regardless of which OS is installed. [...] It's tempting to see this as a bit of a con. They've sped up the front end so it feels like you're getting more done, but in terms of real productivity it's no better than Vista."
Trick Question (Score:5, Insightful)
Is Windows 7 Faster Or Just Smarter?
I don't like either of those options, how about "just more of the same Microsoft software?"
I understand the article points out that they went with simply a "more responsive interface" paradigm (Web 2.0/AJAX, anyone?) and probably didn't really fix any serious problems. But at the same time this headline reeks of either marketing or hilarious lawyer type questions. Examples:
Productivity (Score:5, Insightful)
The productivity would actually increase if the front end speed increased since it would allow the user to interact faster etc. The other tests such as encoding etc are really CPU and application dependent and not very much OS dependent, so it's not really a fair test.
Smarter not harder (Score:4, Insightful)
Personally I'll stick with Homer Simpson's motto: "If something is hard to do, then it is not worth doing." Which is my rule regarding installing new Microsoft Operating Systems.
Just to throw out one more gem; "If it isn't broken it doesn't have enough features yet." Which seems to be Microsoft's golden rule.
Bad benchmarks for productivity. (Score:5, Insightful)
Video encoding is a terrible metric for "productivity" since it's something the computer can do on it's on while you go get tea. It's pretty much CPU and memory bound. The underlying OS shouldn't be doing anything but getting out of the way.
But UI "tricks" are an improvement. If find it easier to start your video encoder, or can do other resource-light things while the video encoder is running at a small cost to the actual encoding speed, then you're making better use of your meat co-processor. Which really is a "productivity" gain.
Okay, but that's still important (Score:5, Insightful)
I really wouldn't expect significantly different scores for something like an office suite or media encoding. Once the OS gives the process all the memory and CPU time it needs, that's basically it. Maybe for games where there could be significant differences in the DirectX flow, but not in general.
But as the article notes, throughput isn't everything. The "up front" speed and how long it takes for a button push to result in action is equally important if not more so. The responsiveness of applications is something an OS can have a significant impact on, and is probably the most important thing for making the computer -feel- fast, and thus giving a better user experience. Hell I've long considered responsiveness to be justification enough for dual-core processors even when a user isn't multi-tasking or running multi-threading apps. So if it's a good enough reason to get a whole second core, it's a good enough reason for an OS upgrade.
It does sound kinda cagey that they're making this one of the main reasons to get 7, rather than improving Vista. But whatever, it's all academic to me.
Worse than that. (Score:5, Insightful)
They don't define "faster" to include the response time of the interface.
But most users DO include the interface response time in their opinion of which is "faster".
I think Microsoft made a big mistake with the "fade in" menus. Just turning them off gives the user the impression that you've made their machine "faster". Even though email works at the same speed as before. As does Word. As do their games.
Re:Bad benchmarks for productivity. (Score:3, Insightful)
Video encoding is a terrible metric for "productivity"
Unless you are encoding it live, straight from the camera.
Productivity originates from the users perceptions (Score:5, Insightful)
I'm no fan of Windows. But improving UI responsiveness, does greatly improve user throughput when using a system - partly because the user can do what they need to do more quickly, but also because there are fewer jarring moments where you are brought out of the process of creation to have to wait on the computer to finish something. These small interruptions can add up to a big loss of focus over a day.
Faster interface = improved productivity (Score:5, Insightful)
A Con! (Score:5, Insightful)
So let me get this straight: Windows 7 is only faster than Vista. It doesn't manage to also make third party programs written for Vista magically faster as well.
AGREED! (Score:3, Insightful)
For the average user, a lot of time is wasted waiting for the UI, or being afraid or unable to do other tasks while something "heavy" is going on (like reading email, surfing, etc.).
If the system still has the same horsepower, but I'm better able to actually multi-task without slogging through a molasses interface, then it's a huge improvement.
It's just not worth trying to type an email sometimes when it takes 6 seconds to update the UI after each keypress... maybe doing so will slow down your build in the background, but only marginally compared to the time wasted if you can't do anything at all during that time.
MadCow.
If it feels faster, you're getting more done (Score:5, Insightful)
They've sped up the front end so it feels like you're getting more done, but in terms of real productivity it's no better than Vista
I take exception to this. Obviously, if the video encoding tests were written well, there will be little speedup. But if a window environment "feels" faster, you actually DO get more done. There is less frustration in waiting, and you can generally multi-task much easier.
There was recently a discussion of a faster X server [slashdot.org]. Frankly, I get plenty done on the old "slow" X server, but if one feels faster, it will actually eliminate a lot of brainpower consumed by waiting on a context switch.
There was recently a discussion on a faster Linux boot-up [slashdot.org], which preloaded your configuration as you're typing your password, and had lots of other fast features... But that doesn't actually speed up Linux, in terms of encoding video. It just makes it "feel" faster.
I like OSS, but I see lots of bad tags being made. Unfair comparisons are simply unfair comparisons. You can't hail a nice feature in one OS, and discount exactly the same feature on a different OS. Without being hypocritical, anyway.
Re:Bad benchmarks for productivity. (Score:3, Insightful)
Video encoding is a bad metric for "productivity" but it is a
very good means to test how well a system will continue to
respond under high load. If transcoding craters your system
then that's a problem. This particular task might not represent
a "productive" part of your normal workload but it's probably
a good stand-in for something that is.
Personally, I like the fact that I can keep my system completely
busy and not be bothered by it. If I have my own "cloud" at home
this means that all machines on the home network can be effectively
utilized for whatever I might want to do.
Re:Productivity (Score:5, Insightful)
Exactly what I was thinking - For most Windows users, the user is a major bottle-neck. By simply responding more quickly to them and allowing them some time to react (even if the system isn't fully ready to react to their next input), you can certainly improve performance. While there are a lot of users that do care about encoding time and Office benchmarks, most users just want IE and Outlook to let them start typing quickly so that they can forward on the latest news regarding Bill Gates paying people for testing their new e-mail system or letting their voice be heard by voting on "Am I Hot or Not?"
It looks just like Vista now because... (Score:2, Insightful)
At this point it probably IS Vista!
Do you think Microsoft re-writes the OS from scratch every time? No, they just incrementally change the previous version, and this happens slowly over the course of development. Since 7 is still a year or so away at this point they're just showing you mostly user interface changes with little or no changes to the core underlying os. By the time it releases there will probably be some significant changes, but right now I suspect you're mostly looking at a UI demo running on top of plain old Vista.
I keep seeing posts like "I tried out Windows 7 and it looks like all my software is 100% compatible and runs great!". Well duh, since it's the same OS you're running now.
Wait six months or so before passing judgement on this thing.
G.
Productivity. (Score:3, Insightful)
Okay here is the big question.
Do many users need a faster PC?
On a clean Windows box when are you waiting on the computer?
I am not talking about games, scientist, or people using CAD/CAM.
I am talking about the average user?
Now when you are waiting how often is it an IO bottle neck?
Waiting for a program to start, waiting for a file to download or some other function like that.
The real answer is that for the most part PCs are quick enough.
Video encoding isn't something that the average users does yet. It will be in the future but right now not so much.
Choice quote (Score:2, Insightful)
What is this knucklehead smoking?
Yeah, the signal it sends is that you're a blithering idiot, a chump, and an easy mark. What a jackass...
Cheers,
A bit smarter would be welcome (Score:4, Insightful)
Improving the front end is overdue and welcome.
Under Windows 2000/XP (have not touched Vista yet) I have often wondered why the Windows Explorer takes ages to show a directory, even if the actual content at the displayed directory level is only a few dozen elements. Maybe it scans all subdirectories for whatever arcane reason?
I strongly suspect there is a lot that can be optimized there, and if Windows 7 finally got around to it, this would be a good thing.
Progress, right? (Score:3, Insightful)
If the UI is now snappier and more responsive so that the user feels more happy with his user experience, isn't that still good progress even if in reality the speedup is only subjective? Everything that makes the user more content using the product is good, right?
Re:Productivity (Score:4, Insightful)
The productivity would actually increase if the front end speed increased since it would allow the user to interact faster etc. The other tests such as encoding etc are really CPU and application dependent and not very much OS dependent, so it's not really a fair test.
Umm encoding is not all CPU and application dependent. Maybe you forgot what an OS does. It schedules when a program executes, where it is located in memory etc. So if Vista puts a program in different places in memory rather then linear or it has a different caching model then windows 7, the execution time will be different. Also if vista does not let the program execute as much as windows 7 the execution time will be different.
Optimizing the UI is perfectly legtimate exercise (Score:5, Insightful)
Of course a snappy UI is a huge deal. Users spend a lot of time navigating before they actually run anything. And, keeping the UI snappy even when the CPU is under heavy load is an especially important user experience requirement.
There's nothing illegitimate or sneaky about optimizing the hardware to better serve the user.
Thank OLPC (Score:3, Insightful)
was much work really needed when consumer computers are quickly becoming powerful enough to actually run vista smoothly?
We can thank Nicholas Negroponte for this. His One Laptop Per Child project inspired the mainstream PC industry to develop similarly low-powered, low-priced subnotebook computers called "netbooks". Windows XP and Ubuntu run better than Windows Vista on the small CPU and small RAM of these computers.
Re:Bad benchmarks for productivity. (Score:5, Insightful)
In which case, as long as it can do it in real time, you haven't got any problems. Making it faster isn't going to help because the bottleneck is the outside world.
I am NOT the kind of guy... (Score:4, Insightful)
That would cut MS any kind of slack. I hate their ugly guts (and boy, all guts are ugly, but theirs...: just imagine winnt's kernel code).
That being said, if the thing is faster in the iface, its a faster experience and that is that.
Those are seconds saved.
Its just stupid to hit them for doing something better, especially if you see what they are coming from: i mean, it cant be that hard to make something feel better than, for christ sakes, VISTA.
Um - A bit early (Score:2, Insightful)
I despise Microsoft, and I doubt the conclusions from the article will change, but lets let them get it past BETA before we burn them as witches.
Where was the real improvement (Score:4, Insightful)
One of the problems with Vista was hardware upgrades. Every new cycle of Windows requires some hardware upgrades for the new version. Unfortunately for MS, the 5 year gap between XP and Vista hurt them. Combined with MS not defining the real requirements of Vista meant that most people trying to upgrade their 5 year old machine would end up in disaster.
These are MS recommended hardware for Vista Ultimate/Business:
Compared to XP Pro requirements:
Now both requirements are really inadequate to use the OS fully. The difference is with only 3 years between 98/XP, it was easy for users to upgrade their CPU, motherboards, video cards without much infrastructure changes. For the 98/XP upgrade it was only 3 years and most users only needed more RAM. If users did require hardware upgrades (CPU, video card), these were readily available. Need a faster Pentium/Athlon in 2001? Go down to BestBuy. The 5 year gap between XP and Vista meant that some hardware upgrades were not easy or even possible. Need a faster Pentium/Athlon in 2007? They don't make them anymore. Ebay is your only real source and even if you upgrade to the fastest one, your system will be slow.
Re:Perhaps this alpha releases uses Vistas kernel? (Score:5, Insightful)
The lipstick on a pig aphorism comes to mind.
W7 is the Vista that Vista could have been. But that may be damning with faint praise.
The sheer obesity of Vista could easily have been improved upon. Somewhere, there is a coder army taking instructions from an idiot. They need to find that idiot and fire that person. Even Gates was better at direction.
Re:Productivity. (Score:4, Insightful)
You're right. However, you're missing an important point: Hardware and software vendors implicitly collude to create a continuous captive market demand for their products.
Windows version "x" won't run acceptably on anything less than a 2GHz processor with 4GB RAM--time to upgrade your computer!
Video card "y" only has drivers available for Vista--time to upgrade your OS!
Support for application "z" has been dropped, and the new version requires more RAM and Windows 7--time to upgrade everything!!!
Honestly, find a modern computer which can run Windows 2000, and you'll have a blazing fast machine. XP isn't _much_ slower, and has the advantage of newer device support.
Strictly speaking, an OS shouldn't have "features" from the user's point of view. Gluing a GUI to the OS was arguably Microsoft's first act of truly evil genius. Same thing with the web browser. THESE ARE NOT OPERATING SYSTEM FUNCTIONS, but they help increase the hardware requirements (and the hardware requirement delta between versions), and hence sell hardware, which sells software, which sells...
In a just world, Microsoft would have taken the code base for Windows 2000, added support for 64-bit multicore processors, newer hardware and so forth, tweaked the UI a bit (XP has some clear advantages--and some clear disadvantages), and LEFT THE REST ALONE! Most of the serious code changes between versions have been for no reason except adding "features" (i.e. stupid crutches and applications), which slow things down.
But hey--it's all about marketing, sales, and profits. That's the reality of the industry.
*and maybe explicitly--who knows what goes on behind closed doors?
Slashdot and Tags (Score:1, Insightful)
Can someone tell me why Slashdot's tags overlap with the summary? I mean do the site maintainers even check their work on more than 1 web browser? It's pretty annoying trying to read the last few sentences of the summary only to have it overlapped with tags. Also what is this metric-fuck-ton of javascript that has started plaguing the front page. I thought this was a news blog not the forefront of web 3.0. Between those 2 fuck ups, Flash ads, and that stupid survey popup, and these new slew of pointless articles (Stupid useless VIM tricks? c'mon) this site is quickly going into the shitty and not even because of comments. Fix your shit slashdot. Please
The front end is what's wrong with Vista anyway (Score:5, Insightful)
If they've sped up the front end consistently, then I would be very happy.
My primary complaint with Vista is how long UI operations take. Opening windows, dragging them around, launching applications etc. all seem to take place in something approximating geologic time.
Once I have a high-performance app open (say a game), the game itself runs pretty quickly. It's the getting there that's a problem.
Re:Worse than that. (Score:5, Insightful)
But most users DO include the interface response time in their opinion of which is "faster"
Indeed, and that's a pet peeve I have with Linux. I use Linux - a lot. Heavily on servers at work (but generally CLI only there), and then at home I have a Linux Mint desktop that I use in addition to my Mac and Windows systems.
I love the concept of OSS, and for someone who when they were growing up saw a compiler as something that cost hundreds of dollars, the whole concept of having such a nice development environment is just amazing.
That said, while actually going from point A to point B probably isn't any slower, the interface just makes the system feel draggy. All the little pauses and and graphical oddities when moving a window around just take their toll, but the actual OS is fine (as obvious when I try to do something like say, compress video or something, where the Linux system holds it's own quite nicely).
Hopefully Wayland will take off and help in that regard. Mac OS X has shown what a slick, responsive UI can do for a Unix-like backend. It just sucks that it's tied down to only a subset of available hardware.
Re:Productivity ... Really? (Score:3, Insightful)
The specs you casually throw out are just astounding.
Granted I don't try to run it on crap systems
To use Dell as the brand, you mean I can't run it on a Dimension/Vostro?? I've GOT to spec Precision boxes?
with less than 2GB of RAM, either.
Granted 64-bit is *the* future, WTF is consuming all those resources? I'd guess it's some DRM/crypto nightmare, but I don't know.
Although my 7 test box only has 1GB of RAM.
Only? I've got a Thinkpad T42 running Debian Lenny and KDE4 will ALL of the eye candy on 512MB RAM with no problems. Disclaimer: 1/2 my mobile work is telnet/serial interface, so my productivity gains are faster CLI-fu and good systems topography.
All of this is to conclusively state that something is seriously wrong at Microsoft when a machine is a dog with only 1GB RAM.
Reviews of Windows 7 are biased (Score:5, Insightful)
Re:Productivity (Score:4, Insightful)
Ok, in that way then Windows 2000 is 100 times more productive than Windows 7.
They will impress me when they get the responsiveness of Windows 2000 on a 2 core modern machine. It's freaking lightning fast.
Re:Worse than that. (Score:5, Insightful)
Can't be too surprised about that - most people are much more concerned about the apparent UI responsiveness than whether they'll shave a few seconds off of a video encode. And given that most people see Vista as very slow and unresponsive, Microsoft would do well to change that perception unless they want to be known for the TWO biggest software disasters in the 21st century.
Re:Mp3 Locking? (Score:3, Insightful)
(w/64 Megs of RAM)
Well there's your problem. I haven't had a computer with 64 MB of RAM since the mid-90s. My phone has 2 GB!
(kidding, of course, I know that post and was hoping that would show up in this thread. However, if you're going to update it for the context, go the full monty!)
Re:Perhaps this alpha releases uses Vistas kernel? (Score:1, Insightful)
As is the way with MS , they update all the eye candy first to get the drooling masses interested , then they get down to the core stuff where it really matters later on - ie the exact opposite way round to the way it should be done.
This is why Linux fails as an alternative. Sure it runs great under the hood but the GUIs are to intrusive for the novice computer user to learn.
There is a reason Windows sells well. The interface is easily learned by the masses.
Re:Productivity (Score:5, Insightful)
This is very true, but a slow UI is what most people will complain about. If someone fires up handbrake, sees two passes of h264 encoding with 30min+ remaining per pass (and that's what I see on my 8-core/10GB system, so most people will be looking at 2-4x that), they'll put that down to it being a slow application. If they go to click a menu item in Handbrake and there's a perceptible delay, they'll blame the OS.
Is either bit of blame entirely fair or correctly placed? Nope. But that won't stop 99% of computer users.
Re:Productivity (Score:3, Insightful)
A fair test would exercise the paging and memory system more. Maybe some interprocess communication. Such as having several active applications running, foreground and background, and seeing how fast they get their job done or you can switch between them. Get a long running compiler build in the background (Visual Studio for that mix of computation and visual fluff and memory bloat) and Word in the foreground, Firefox, an Excel spreadsheet, Outlook, Matlab, etc.
Re:Worse than that. (Score:5, Insightful)
Re:Mp3 Locking? (Score:5, Insightful)
So ... the Better than Average Edition is the basic edition, right?
It's like the popcorn sizes in the movies. Now they're called large, extra large and super size. Funny enough, they're just the same size the old small, medium and large sizes. Only the price changed.
If Microsoft really wanted to improve productivity (Score:3, Insightful)
Re:Bad benchmarks for productivity. (Score:3, Insightful)
Re:Worse than that. (Score:5, Insightful)
Having worked in marketing and well as puely technical roles it is clear as a bell what is going on here.
When new product uptake isn't up to projections the marketing dept. has a few options. One of the options in its arsenal is to "relaunch". Windows 7 is clearly a "relaunch" of Vista. With all the development time and Money put into Vista don't think for a second that they can develop yet another code base in a fraction of the time. It is the same product with a different name.
Relaunches are used when there is a perceived problem in the marketplace and the engineering dept.says the product is sound.
Re:Where was the real improvement (Score:1, Insightful)
Re:Trick Question (Score:5, Insightful)
Does anybody else remember when the greatest thing in Windows was After Dark, with it's screensaver of flying toasters? what we really need now is a repeat of that, but with chairs instead.
Re:Worse than that. (Score:1, Insightful)
And yes Macs cost more and are of higher quality than bargain bin PCs.
How exactly are they of higher quality? Do they somehow get better quality CPU's from Intel off the production line? Better quality RAM chips? Hard drives? GPU's? I'm curious to know.
Re:Productivity originates from the users percepti (Score:2, Insightful)