Slashdot Mirror
Windows 7 Benchmarks Show Little Improvement On Vista
snydeq writes "InfoWorld's Randall Kennedy examines Windows 7 from the kernel up, subjecting the 'pre-beta' to a battery of benchmarks to find any signs that the OS will be faster, more responsive, and less resource-intensive than the bloated Vista, as Microsoft suggests. Identical thread counts at the kernel level suggest to Kennedy that Windows 7 is a 'minor point-type of release, as opposed to a major update or rewrite.' Memory footprint for the kernel proved eerily similar to that of Vista as well. 'In fact, as I worked my way through the process lists of the two operating systems, I was struck by the extent of the similarities,' Kennedy writes, before discussing the results of a nine-way workload test scenario he performed on Windows 7 — the same scenario that showed Vista was 40 percent slower than Windows XP. 'In a nutshell, Windows 7 M3 is a virtual twin of Vista when it comes to performance,' Kennedy concludes. 'In other words, Microsoft's follow-up to its most unpopular OS release since Windows Me threatens to deliver zero measurable performance benefits while introducing new and potentially crippling compatibility issues.'"
What exactly is this article trying to prove?
Microsoft themselves have said that Windows 7 will ship will the same underlying infrastructure as Windows Vista. They also said that Windows Vista was the biggest kernel rewrite since Windows 2000.
The interesting thing about a lot of Vista's bloat is that it isn't kernel level. We know this since we can compare Windows 2003 and Vista. Windows 2003 has almost identical program startup times to Windows XP/2000.
I do think that Windows 7 is going in a disappointing direction in general. They seem to be playing right into what I like to call the "Apple Trap." Instead of doing what Microsoft do best which is to produce a workhorse they instead try and play the designer, and want to make a work of art.
"general" - is that another word for zero? - because I have yet to see a business running Vista, and I certainly don't think they are running Windows 7 - or probably ever will be.
Genesis 1:32 And God typed
I have no problems with benchmarking the O/S and commenting on performance and the like, but when the person that analyzes and presents these results says: "the process lists are similar" I'm forced to wonder what the guy is smoking. OK, so you have have smss.exe, csrss.exe, winlogon.exe, a bunch of svchost.exe processes. That really says nothing about the underlying architecture of the operating system and the amount of differences that are there. This guy might as well have said "I looked at Word '97 and Word 2007 and they're both named 'winword.exe' and let you edit text. I'm struck by those similarities!" Anyone expecting Windows 7 to be a radical departure from Windows Vista is delusional, all the more so if that expectation involved vastly different process lists. Also, this guy compares the video encoding performance of Vista and Windows 7 and says there's no performance improvements... That has got to the dumbest thing I have ever heard. Seriously. It might very well be that Windows 7 is as slow as Vista. Maybe it's even slower. But you will never know that by comparing how long video encoding takes on each of them. Video encoding is a CPU-bound process, so nothing Windows 7 does can improve the video encoding performance of any machine because it cannot just magically improve your CPUs clock speed. All other things being equal, any gains from encoding german scheisse porn on Windows 7 over doing so on Windows Vista are going to be negligible at best.
*cofff* Ubuntu 8.10 *cofff*
So we're skipping this one as well?
One of the biggest PR failures of Vista was serious compatability issues with old software and hardware. (I'm going to blame the soft/hardware makers for this. Everyone had 5 years to collect an arsenal of XP gear so I don't think they cried themselves to sleep that we had to buy new Vista Compatible printers just because they couldn't be bothered fixing the drivers.) MS have decided to base Win7 almost entirely around the existing Vista kernel to avoid this, hence the identical performance. "[I]ntroducing new and potentially crippling compatibility issues" would be more likely if MS had decided to chase performance improvements in Win7, unless they based Win7 around the old XP kernel (which ain't happening in their new one-kernel-to-rule-them-all approach).
No kidding!!! What do you say at this point?
Ah, but the other article was about how Windows performed the same computationally while having a faster interface. That failed to needlessly bash Microsoft by extrapolating miles from the evidence, and therefore was insufficient.
No kidding!!! What do you say at this point?
He said "better" not "more".
Quality not quantity, sadly.
I love Linux, it's good for work, good for application and data servers, but for me, there is a problem.
I am a gamer and I like trying out new hardware. Both of these always pose problems under Linux.
Good stable drivers take time, and require the support of hardware vendors.
Sadly, this means I still have to own and use a copy of Windows XP or give up on games and toys. Ain't gonna happen!
Please just stop following every step of "Windows 7" which will probably not be out for years, despite anything Microsoft says.
The only thing those reports generate is the hype Microsofts wants around their unreleased OS to keep up hope in people dissatisfied by Vista. Yeah, this time it's all going to be better...sure.
Windows 7 is not special and it's not worth reporting every tidbit unless there's actually a product or a set-in-stone feature list.
The current release isn't a release candidate. It's not a beta. It's a PRE-beta. Microsoft have about at least 10 more months until they call Windows 7 done.
Steven Sinofsky specifically said in his PDC 2008 keynote: "please don't consider this build suitable for benchmarks", but does anyone listen? Nah, let's run the benchmarks! :)
Okay, this has bothered me for over a decade.
What makes anyone think that the next release of an OS is going to be faster? It's not going to be. I don't care who developed it, either, whether it be the giant of Redmond, the hipster of Cupertino, or a bunch of unwashed shut-ins writing lines of code in their moms' basements. Every iteration of an OS is actually going to be slower, and that is just a consequence of it doing more.
The only real question, then, is if the balance between the added functionality and the slowdown is coming down enough on the functionality side to stop people from getting pissed off. For XP, the balance was nice. For Vista, it's not. For Tiger, it was. For Leopard, I guess it's not for some people (but it is for me). Linux doesn't do anything regardless of distro or update, so it's kind of hard to talk about.
The point of the story is this: I don't actually care if something doesn't run that fast, because I'll probably replace my hardware before that OS runs its course, and it'll work great on the next kit. All I really care about is if it runs well enough to enjoy the added benefits of that extra code.
Microsoft's obsession with backward compatibility is killing it.
For home and gaming, they need to keep XP and disable it from being used in a business network... let that horse run as far as it can.
For business and other work, they need to write a brand new kernel and everything and start over learning from all previous mistakes and discarding backward compatibility... natively. Then build a VM compatibility layer with the intent that people will use it in the process of weaning themselves from Win32 and all that backward compatibility and supporting broken applications nonsense.
Been saying this for a long time and will keep saying it. I said this before Mac OS X was announced. Apple, it would seem, had the same idea and it is working VERY well for them. The compatibility VM sucked bad which actually prompted people to upgrade their apps even faster. And no one stopped using Apple over it. And no one stopped developing software for Apple computers over it. It was a burden on users and developers to make that change, but in the end it was the best move.
Microsoft is another story. When you are in control of everything, that is precisely what you stand to lose. But ultimately, I see things are coming to a head and Apple sees it too. No matter what Microsoft does, they will lose. They need to make plans to limit their loses and plan for the future -- not just two years of profit forecasting.
If W7 can do the compatibility part right here, it's a good thing, not a reason to look down on it for not being different enough. How typical of Slashdot -- would you honestly ever be able to use the same logic about your favorite OS?
Beware: In C++, your friends can see your privates!
Are you comparing Vista to XP, DX9 to DX10, or your graphics card's Vista drivers to its XP drivers?
(hint: it's a mix of all 3, but the last will make by far the most difference in a graphically-bound game)
I couldn't agree more.
Linux is a great O/S and makes wonderful servers but as a desktop it just doesn't have the software. I earn my living using Photoshop, Cubase, Sound Forge and CD Architect.
I really couldn't care less what O/S my desktop runs just as long as I can get my work done. Sadly the Linux equivalents don't yet cut the mustard so I too am stuck with XP.
Linux is a great operating system which is only missing professional desktop applications.
That's funny; I though the article was about the operating system and not the market that was created around it. In terms of "support", Linux is far better than Windows. It's the support of the environment of the OS that's the problem. Linux supports gaming. Only the games hardly "support Linux". Devices are very well supported on Linux. Windows does hardly support any device at all out-of-the-box. It's the drivers that you get on the friggin CDs (where do I get a USB CD drive these days?) or downloadable from a website that add the support. Office software (I suppose the only office software you know of is Microsoft's office) all support nearly any OS including Windows. If you consider all office software, support is quite nicely spread.
There are loads and loads of scientists/students who still prefer [La]TeX to a graphical word processor any day. There's something about expressing your ideas straight away in a fast and light editor, and producing professional quality documents without any graphical tweaks, rather than wasting memory and processing power for a glorified Paint while praying it not to crash.
Escher was the first MC and Giger invented the HR department.
Yes good, efficient design is from the ground up but once you've got the underlying structure sorted, you then move onto features. Once all the features are in place, then you move onto optimisation. Optimisation and bug fixing are the final stages of development, after all, you can't optimise things which haven't been implemented yet can you? Often, yes, you do get magical performance boosts late into development (have a look at videogame development for the clearest examples of this). However Microsoft have never promised magical performance boosts. They've just said less bloat, more streamlining.
No new features? There's the improved wireless, the GUI which will now load and be smooth BEFORE graphics drivers are installed (I don't believe any desktop versions of windows have done that since before win95), the interface is hugely optimised, resulting in a much smoother experience from practically everyone who has done the beta. They've shown a version that will run comfortably on netbooks whilst still looking and feeling great (and the OS is SSD optimised). They improved the UAC so you can make it as invasive or as invisible as you wish. They've implemented Libraries, Homegroups, a 'Play To' feature that will let you play media on any connected PCs. They've updated all the basic applications (notepad etc.).
Do you really think that counting threads and memory footprint will give you any sort of indication of a systems performance? So, whatever those threads are really doing is not useful information? By design Windows uses as much memory is available, as unused memory is of no value. A performance indication would be to measure how much actual pagin is there when physical memory is exhausted by running process. Counting used memory is worthless. And counting threads and processes? Come on! What sort of analysis is this? Even if it were based on the final product (instead of a pre beta version), this analysis doesn't tell absolutely nothing. Not that I would expect that Win7 uses fewer resources that Vista. It would be a great thing if, coming a few years later, it used the same level of resources (meaning it should be able to run in machines over five years old) but expecting it to consume fewer resources is delusional. Performance today has much less to do with resource usage than with responsiveness and proactivity anyway.