Forgot your password?
typodupeerror
PC Games (Games) Operating Systems Software Windows Entertainment Games

Windows 7 Gaming Performance Tested 179

Posted by Soulskill
from the take-with-a-beta-grain-of-salt dept.
Timmus writes "Gamers holding onto Windows XP may not have to fear sluggish performance when Windows 7 debuts. While Windows Vista's gaming performance was pretty spotty at launch, the Windows 7 beta build seems to handle most games well. Firingsquad has tested the Windows 7 beta against Windows XP SP3 and Vista SP1 on midrange and high-end gaming PCs across 7 different games. While the beta stumbles in a couple of cases, overall it performs within a few percentage points of Windows XP, actually outrunning XP in multiple benchmarks."
This discussion has been archived. No new comments can be posted.

Windows 7 Gaming Performance Tested

Comments Filter:
  • by AaronLawrence (600990) * on Tuesday January 27, 2009 @07:56AM (#26620339)

    Their benchmarks hardly show a conclusive improvement for Windows 7. Vista mostly beats it in DX10, and XP still beats it about half the other benchmarks. It *does* manage to beat Vista in DX9... hardly exciting, but something.

    Their mid-range also seems a bit ambitious - more like mid-range of new hardware for serious gamers, which means high-end for the rest of us.

    The most interesting paragraph for me:
    "because Windows 7 felt more ready to go once the desktop loaded up. Both XP and Vista took at least an extra minute after the desktop loaded to be ready to run applications, while Windows 7 ran Firefox without stuttering or hesitation. "
    Now thats something worthwhile. The 2 seconds of "boot time" is irrelevant, being able to use the desktop immediately is a real improvement.

    • by ozmanjusri (601766) <(aussie_bob) (at) (hotmail.com)> on Tuesday January 27, 2009 @08:39AM (#26620611) Journal
      Their benchmarks hardly show a conclusive improvement for Windows 7.

      They don't show an improvement at all.

      In most of the tests, even Vista is faster, and in the few where Windows 7 wins, it's by so little it could be within the margin of error of the tests. As the article says, the differences are most likely driver related rather than intrinsic to the OS.

      And the reason they're so close is that Windows 7 IS Vista with a few tweaks and a hell of a lot better marketing.

      • by Jugalator (259273)

        Also note that this is a beta against RTM releases.

        It could play a role.

        • More importantly, they're going to be using beta drivers. Remember when Vista first came out? The drivers were a joke. Half of the crashes were attributed to nVidia. To me this says more about the drivers being used to run the tests than the OS's themselves.
        • Windows 7 isn't much of a beta. It's just a (big) service pack for Vista. There are some significant changes but not enough to really be a new OS. That's why it runs quite well as a "beta."
    • No Wine? (Score:3, Interesting)

      by Ed Avis (5917)

      Disappointing that they did not test performance on Linux with Wine or Crossover Games. Not every game will run on that but for those that do, the performance comparison could be very interesting. They could also test the performance of the games under ReactOS [reactos.org]. Comparing several releases all from the same company, always from the same one company, gets boring after a while.

      • Well, I like Linux and all, but when it comes to gaming articles published on a big gaming site, posting Linux benchmarks would be a bit like comparing a train's performance on Russian versus European rail grades, interesting but kind of pointless.

        I know a lot of people are holding on to Windows software for gaming (like myself, I would switch if it was easy to run all my games) and Linux has potential, but game publishers just aren't interested in such a financially insignificant market. They are the ones

      • by ianare (1132971)
        My limited experience is based on the only game that keeps my NTFS partition populated : Civilization 4. On a newer machine : 4GB RAM, ATI 4850, intel Q6600, the game is simply unplayable in wine 1.0.13, but hauls much ass in XP. Verdict - dual booting is still in my (immediate) future.

        Just my $0.02 ...
    • Re: (Score:3, Insightful)

      by Spatial (1235392)

      "because Windows 7 felt more ready to go once the desktop loaded up. Both XP and Vista took at least an extra minute after the desktop loaded to be ready to run applications, while Windows 7 ran Firefox without stuttering or hesitation. "

      That's a bit weird. My PC with a 14-month-old XP SP2 install is running applications perfectly less than five seconds after I log in.

      My old P4 with a six-year-old install took maybe 5-10 seconds at a high estimate, with a total boot time of less than 25 seconds. What have they got on this thing that it takes such an incredibly long time? A minute is something I might expect out on a crap laptop running Vista. For XP on a modern desktop it's crazy.

  • Printable version (Score:5, Informative)

    by L4t3r4lu5 (1216702) on Tuesday January 27, 2009 @08:15AM (#26620449)
  • 64bit or 32bit? (Score:5, Interesting)

    by CaptainNerdCave (982411) on Tuesday January 27, 2009 @08:19AM (#26620479)

    the story doesn't mention, but this is key.

    first, they compare a 32bit xp to a 64bit vista; oranges to grapefruits.
    next, they add windows 7 and don't mention if it's 32 or 64.
    they did a decent job of being objective... but still fell short of offering us the information that we need. does 7 implement 32 and 64bit functionalities as smoothly as vista64? is it the kind of angry child that 64bit xp is?
    bad grammar aside, this review is lacking some fundamental information that should have been disclosed on the first page.

    • Re:64bit or 32bit? (Score:4, Informative)

      by kitgerrits (1034262) * on Tuesday January 27, 2009 @08:40AM (#26620619)

      Keep in mind that XP x64 is not actually XP, but Win2003 x64 with some changes to make it look and feel like XP.

      • by rwa2 (4391) *

        Yep, exact same kernel and everything. On XP x64, I hear there's even a properties page somewhere that still says "Windows 2003 Server x64" (maybe it was even the sidebar of the start menu).

        I wouldn't know firsthand, since I actually run Windows 2003 Server x64 on my laptop.

        http://www.osnews.com/story/3655 [osnews.com]

        It's OK, but I should disable the comment I'm forced to write for the log every time I want to shutdown or reboot.

    • Re: (Score:3, Interesting)

      by jellomizer (103300)

      Most people don't realize this but 64bit is slower then 32 bit. Every instruction is twice as long, Executables are larger, and more IO back and forth from CPU to memory.

      This isn't realized because of the speed improvement from 16bit to 32bit computers back in the late 80's/early 90's. Because by then most 16bit Systems built and sold had Maximum Ram 640k already installed, and actually stalled development of systems with more Ram for a while. When the 32bit systems came out they had 1 Megabyte. Which was

      • Re:64bit or 32bit? (Score:4, Insightful)

        by ozphx (1061292) on Tuesday January 27, 2009 @10:31AM (#26621739) Homepage

        Most people don't realize this but 64bit is slower then 32 bit. Every instruction is twice as long, Executables are larger, and more IO back and forth from CPU to memory.

        Your FSB is fast enough to deal. The pipeline in your CPU works on bigger chunks anyway. Heck, thats why vector processors were invented, MMX, SSE, etc - registers weren't wide enough. Executable size is insignificant, and instructions are varying length anyway (opcodes don't suddenly become 64bit on x64).

        Heck they are not even properly using duel core.

        Theres almost a thousand threads on my current box. I'd say thats taking advantage of both cores. Apart from algorithms that flat out don't parellelize (eg MD5ing a bunch of data) it seems to be going pretty good here...

      • Most people don't realize this but 64bit is slower then 32 bit.

        Interesting theory.

        It doesn't seem to be borne out by benchmarks [phoronix.com] though, at least for Linux.

      • Of course, they aren't twice as slow. (i.e. They don't run at half the speed, for the pedants.)

        The slowdown is a few percent, because most applications aren't bound by memory speed of instruction fetching.

        But who really cares if the OS is a few percent slower? It uses ~0% of my cpu anyway. Most applications I run are still 32-bit, even though my OS is XP x64.

      • Re: (Score:3, Informative)

        by rwa2 (4391) *

        I've done some quick benchmarks on identical hardware using PerformanceTest.

        Yeah, 64-bit executables are a bit bigger than 32-bit executables, so they take marginally longer to load from the disk. That's probably why you're confused.

        But looking at the benchmarks, integer operations are much faster (2-3x), floating point and memory operations are a bit faster (10-20%), and disk access is marginally faster (5%-10%). There was no difference on memory writes for some reason.

        There was no difference on 2D or 3D

      • The "64 is slower than 32" argument works for every single (32/64-bit) CPU architecture out there... Except the one most relevant to this article.

        The 64-bit extension to "x86" (wether you call it x86-64, x64, AMD64, EM64T, or Intel64,) doubles the number of registers, as well as implementing a brand-new, mandatory FP unit. Therefore, "generic" 64-bit code on x64 will run better than "generic" 32-bit code on the same processor. By a measured increase of 10-15% in most benchmarks. Because Windows is compl

        • by Firethorn (177587)

          Second, even if your workload consists of apps that are purely single threaded, you still benefit from two cores by letting your workload run on one core, and everything else on the second.

          The single biggest thing I noticed when I built my first dual core computer over my older single core one. Didn't even notice the graphics improvement that much.

          As a result, I'm unlikely to ever go back to single core. I figure quad core will be my next upgrade; but I bought a dual for the time because metrics had dual > quad > single for the price of the CPU for users of mostly single threaded applications, even many multi-threaded ones. The simple reason was that, for the price, quads took a good

      • While it is true that a 32-bit app usually runs a bit slower on a 64-bit system than a 32-bit system (we are talking a few percent) I think it is as much or more to do with the small amount of overhead the Windows on Windows emulation incurs rather than larger words. The FSB has plenty of bandwidth for that.

        However what you discover is that when apps are recompiled to 64-bit, they are often faster, even if they shouldn't be. Prime95 found that. The program is all FP calculations and is done using SSE2 instr

      • by canajin56 (660655)

        If you say so. Comparing 32 and 64 bit XP, 32 bit game benchmarks are near identical on my system, well within margin of error. 64 bit capable games, such has HL2, have a notable increase in speed beyond the margin of error. The reason is several fold.

        "Every instruction is twice as long." Nope, opcodes and their paramters have always been variable length. In x86, they aren't always exactly 4 bytes. In x64 they aren't 8 bytes long, either! In x86, you'd have a function call CALL DWORD PTR [ADDRESS]

    • doesnt matter (Score:2, Insightful)

      by cefek (148764)

      Running Vista64 since day one, and Windows 7 for a while, I must say internally both systems look similiar. Some article TL to quote stated that Seven is to Vista like 98SE was to 98. It does not take a rocket scientist to guess MSFT would never release such a dud like XP64 again - it's been overdone (can you say that?) by now.

      • by Skal Tura (595728)

        XP 64 ain't a dud, it's a necessity of life for some of us who actually wants newer than 3yr old hardware, and take advantage of the hardware.

        The maximum limit of ram is 3-3.5Gb on 32Bit .... And i'm maxing 4Gb daily on REGULAR web dev utilization.

        Never mind all apps sucking more and more each year. Firefox takes EASILY 250megs of the pie.

        Think of: Firefox, Chrome, IE, Eclipse (or Zend Studio), WinSCP, 5 instances of Putty, Messenger, Outlook, Excel, F-Secure (And some other bg apps) all running at the same

    • Re: (Score:2, Insightful)

      by sponga (739683)

      The whole article/title just reeks of wrongness and is too early to test.

      Testing games on a beta for performance?

      First off any real online PC Gamer who hasn't done it already has found out that Punkbuster doesn't exactly work online for Win7 without tricking the Services options. Not news since this was the story on Vista, only story is the customers demanding the developers to get some drivers and compatibility.

      The drivers aren't exactly out or have not been polished, so you will still have some performanc

  • While the beta stumbles in a couple of cases, overall it performs within a few percentage points of Windows XP, actually outrunning XP in multiple benchmarks.

    and Windows XP out runs windows 7 and Vista in other benchmarks too..

  • by Anonymous Coward
    From the abstract: "overall it performs within a few percentage points of Windows XP, actually outrunning XP in multiple benchmarks". Windows XP was released in late 2001, and in the almost-8 years since, Microsoft has managed to improve performance for my games by "a few percentage points". Not, "alot of percentage points", but only a few. If XP is only marginally worse off then Windows 7 will be, then whats microsoft working on? The flashy looks that I dont see when I have a full screen of zombies being
    • Re: (Score:2, Interesting)

      by maxwell demon (590494)

      It can mean a lot of things:

      • It might mean that XP is already so close to the theoretical maximum performance that it's impossible to improve much.
      • It might mean that the basic architecture of the Windows operating system doesn't allow much improvement, i.e. you'd need a different OS design to get better.
      • It might mean that Microsoft just didn't work on improving game performance.
      • It might mean that in the tested configuration the bottleneck wasn't Windows.
      • ...
      • by Joce640k (829181)

        In a game the OS doesn't do very much, it's all game code vs. graphics drivers.

        All the OS is doing is basic housework like reading the keyboard and mouse.

        Sound is a tiny percentage of CPU power these days and is probably mooted by a multicore CPU.

      • Yes. Since a game is almost the canonical exclusive mode task, the OS should add as little overhead as possible, with as many cycles as possible going the the CPU and GPU.

        If one version of Windows had a lot higher performance than another for gaming, that'd be more suggestive of a bug or driver issues in the slower version.

    • by Sockatume (732728)
      Managing to come up with an OS with radical changes designed to support novel new features (Win7's multicore support is apparently a big improvement) and do so without increased overhead, and actually improving performance, is quite an achievement. New OS releases aren't normally intended to improve performance on a particular system. As an example Ubuntu's a couple of decades of development ahead of Amiga Workbench, but I don't grude that it has system requirements approximately one hundred times higher.
  • EDITORS: EDIT! (Score:5, Insightful)

    by argStyopa (232550) on Tuesday January 27, 2009 @08:40AM (#26620613) Journal

    OK, FFS can we stop linking to the BULLSHIT 16 paragraph=16 page articles that are meant to maximize web traffic? PLEASE?

    Jesus, please: just copy the damn printable link and get it all on one page.

    Slashdot is a fairly heavy-traffic site. You have the throw weight discourage this HORRIBLE style of web page design.

    If the print-summary page isn't available then link the CONCLUSIONS page...readers who are smart enough to parse what WinMark scores are can *probably* figure out how to get back to the detail pages.

    Here's the damn link: http://www.firingsquad.com/print_article.asp?current_section=Hardware&fs_article_id=2404 [firingsquad.com]

    • by mgblst (80109)

      How dare anybody be rewarded for their hard work, which i really want to read.

      Maybe you go into work everyday and don't expect to get paid, but some people like getting money.

  • Waste of time. (Score:5, Interesting)

    by ledow (319597) on Tuesday January 27, 2009 @08:52AM (#26620709) Homepage

    8 years. 8 bloody years. 8 YEARS. EIGHT... YEARS. Say it to yourself.

    What the bloody hell has MS been doing for the last EIGHT YEARS? XP *still* outperforms their only other two Microsoft offerings in the market since its release. In the eight years BEFORE XP, we start with MS DOS 5.0 and Windows 3.1 (remember those days?), go through Windows 95, 98, most decent versions of NT and then Windows 2000. From them to XP... spot the difference. Now jump forward eight years instead and look at the difference, eight years on from XP and what have we got? Next to nothing. Oh, a couple of XP Service Packs that made more difference than every *OS* they've released since.

    I looked at every graph on the page and they are all within a reasonable margin of error, especially in the absence of certain details (i.e. are the drivers all optimised for XP, Vista and Windows 7 equally? Was Windows 7 running 32- or 64-bit? etc.). There's nothing there that'll make gamer's go "OOHhh... gotta have that". It's more like "Well, if I do get lumped with Windows 7, hopefully it won't be much worse than my existing, well-configured, XP install".

    What the hell have they been doing? I've argued before that there are no significant, new features in Vista and/or Windows 7, a myriad of problems still exist with both (and with XP for that matter), the minimum hardware is increasing all the time just to do the same tasks and there's no performance improvement at all (in fact, with Vista, it's quite likely to be the opposite depending on your uses/hardware). They haven't even bothered to comply with most of the legal demands on them in that time. They sort-of-but-not-quite started documenting SMB/CIFS, which hardly kills your current development teams. Is the code for Windows *really* that bad that this is all they could manage?

    Alpha, beta, fine - I expect it to be flaky. In fact, I expect all sorts of debugging code and slagging the disk to death while it churns through buckets of debugging data so they can actually fix real-world problems. However, it builds on Vista drivers which, despite much fuss, are pretty well established now. It performs *identically* to Vista in a lot of tests (which suggests that not much at all has changed under the hood, as does the fact that Vista drivers are still compatible). The new features are basically plug-ins to the existing systems, not massive rewrites of critical code. This all leads me to believe that Windows 7 is a Vista Service Pack, to all intents and purposes. So what the hell were they working on for those 8 years of development with one of the largest software development teams in the world?

    • by Anonymous Coward

      The general sentiment and tone of your comment is exactly the same of when XP was getting ready to come out. Everyone at Slashdot swore left and right that XP was bloated to hell, that it'd run too slow, and nobody would buy it, and it would signal the downfall of Microsoft. The interface so horrible that Joe Sixpack was finally going to wake up and switch to Linux. If not the interface, then product activation would. No, Win2k was their last great OS, and it can't possibly get any better than that.

      And, her

      • by ledow (319597)

        "The general sentiment and tone of your comment is exactly the same of when XP was getting ready to come out."

        Not for me. 2000/XP was a quantum leap. I didn't think it was a good one at the time (mainly because of system requirements) but it changed an awful lot. And it did it in such a way that it was soon on every computer. That happened in a handful of years too. I'm not so sure I could have said that the jump itself from 2000 to XP was so massive - I still get people who bring me laptops with 2000

        • by Yunzil (181064)

          Because... XP-> Vista is hardly a leap at all, it hardly compares to some of the minor 9x updates that occurred, but because it was parts of the GUI that changed, people think that's somehow more miraculous. There was also significant breakage for very little reward. Suddenly, everything needed new drivers to be rewritten, which often meant new hardware, or unsupported configurations.

          Read this paragraph to yourself again. XP>Vista is hardly a leap at all... except for an entirely new (and better) dri

          • by ledow (319597)

            So I've got to download, install and test new drivers to make all the hardware that always used to work continue to work, in the process obsoleting quite a lot of perfectly good hardware into the bargain because nobody (read: manufacturers who have financial incentive to release "Vista" versions of the hardware) can be bothered to make a Vista driver. And Vista drivers provide what advantage? Possibly better security with some extremely crap drivers but I haven't seen evidence of that. Significant breaka

      • by mgblst (80109)

        The general sentiment and tone of your comment is exactly the same of when XP was getting ready to come out.

        That just makes it even worse. When I did use XP, I changed the style to classic, so the major change in XP was to make it look childish shit. All the time, Mac OS was looking amazing. Win 2k is still their best OS, because XP IS Win 2k, with a new shitty interface, more bloat, and some bug-fixes. Congrats Microsoft.

      • by jpmorgan (517966)

        And when 2K was release, /. was giddy over the leaked '24,000 bugs' memo. Everybody was talking about how it was yet another crappy OS from Microsoft.

    • Re: (Score:2, Insightful)

      by Renozhin (1423301)

      Operating systems are UIs; they are not intended to be performance boosters. Their task is to deliver the information that you need in a timely, neat and conclusive manner. As far as I hear, Windows 7 does this better than Vista as well as (at least when it comes to the neat part) XP.

      Try thinking of it in terms of a human body: we're all pretty much the same, but with some differences in component performance. It all comes down to how the body is used and maintained.

      Go ahead. Make a sex joke.

      • Operating systems are UIs; they are not intended to be performance boosters.

        No... operating systems are not UIs, they are only a part of it (sometimes). Aero exists as just a subsystem in Vista, and you don't need anything but the command line for Unix.

        Operating systems exist so that the applications I use can be run on the hardware I own. That's really it.

      • by ledow (319597)

        UURRRK. Operating systems are UI's? No they aren't. They are operating systems. This is the problem - MS has tied the GUI to the OS and now people associate the GUI with the OS. The OS is several layers below anything that needs to draw on the screen, always has been, still is. To make your OS boot even depend on there being graphics hardware is just a ludicrous assertion. This sort of thinking only leads to the statements I hear along the lines of "Oh, it looks different, it must work much better."

    • by brkello (642429)
      And newer versions of Linux are slower than ones from 5 years ago. Where have you been? As computers get more memory and processing power, OSs are updated with more features to take advantage of them. Microsoft decided to incorporate more security and features and this slightly slows down game performance. Your idea that new OSs are supposed to speed everything up is false. That has never been the case for any OS unless it was specifically designed to be fast and light rather than a full featured OS.
      • You really can't beat some of the old, light weight OSes for speed. DOS will probably give you the best app performance you'll ever get. Why? Because DOS doesn't do anything. DOS will load your program and then get the hell out of the way unless you specifically place a call to one of its few services. It doesn't manage memory, it doesn't handle processes, it just does disk operations (it's well named).

        For that very reason, you still see it used in embedded systems today. It loads up whatever app it is that

        • You really can't beat some of the old, light weight OSes for speed. DOS will probably give you the best app performance you'll ever get. Why? Because DOS doesn't do anything. DOS will load your program and then get the hell out of the way unless you specifically place a call to one of its few services. It doesn't manage memory, it doesn't handle processes, it just does disk operations (it's well named).

          The downside, of course, being keeping ems, xms, qemu and what not straight; manually tweaking the load or

    • by master_p (608214)

      Please sir. You do not understand. Windows ain't done till Duke Nukem Forever can run!!!

    • by Skal Tura (595728)

      Vista is actually way more responsive than XP in most cases, it's the reaping every fraction of a percent of performance out of the system where Vista lacks, which is quite understandable given the responsiveness benefits Vista has.

      That being said, i use Both Vista Professional 64bit and XP Professional 64bit daily. Vista at work, XP at home.

  • A new version of an operating system actually working better and more efficiently than the last version? Oh my how will this end; surely someone have made a mistake. My past experience dictates that this is exactly the opposite of how things should be. Each new version should siphon away more and more of the resources to pointless and trivial tasks to ensure that we never get the full capacity of our computer; thus increasing our incentive to buy new shiny gadgets and hardware! Oh bring back the glory days
    • by Narpak (961733)
      And to think it only took them eight years and one eight-billion-dollar miss-try to do it slightly better.
  • For me, the single biggest problem with Windows 7 gaming is the lack of PunkBuster support, as EvenBalance are refusing to support the beta at this stage.

    Fair enough, it is their choice, however the beta is public, so many of their customers are in the same boat here. It seems sad that the public statements seem to indicate they are not even willing to look at it.

    • by mobby_6kl (668092)

      It may not be supported, but at least PunkBuster works on 64 bit Win7, at last for come definitions of "work". The first time I launched ET:QW I got kicked within seconds for not having PB installed. I used their update utility to install it, but still got kicked out of the game. Turned out the PB service didn't install properly or failed to start.

      Searching the net suggested that it's pretty much impossible to get PB to work on Win7, and especially the 64 bit version. However, randomly dicking around with c

  • Missing test: OpenGL (Score:3, Interesting)

    by pak9rabid (1011935) on Tuesday January 27, 2009 @10:53AM (#26622083)
    Although it is the minority API used for gaming, it still does exist. As long as John Carmack [wikipedia.org] is still pumping out gaming engines, there will be games based on OpenGL. Does anyone here have any first-hand experience on OpenGL performance in Windows7?
  • Not when I try to run X3 Terran Conflict.

    Because the Tages DRM garbage that the publishers insisted on inflicting on Egosofts game isn't compatabile with Windows 7 even though it works OK on Vista (which came as a shock).

    Are Tages doing anything about it ? Nope, they take the Creative Labs view that they don't support beta systems.

    I leave it to the reader to decide whether the defective by design belongs with Tages or Microsoft.
  • Jesus people (Score:2, Insightful)

    by ParanoiaBOTS (903635)
    Let me preface this comment by stating that I am not a MS fanboy by any means. But I do have to say this about Windows 7: Yea so Windows 7 isn't as fast as XP. Did anyone ever really consider the fact that it is a newer OS that is doing MORE than XP? The fact that it looks better, and does more than XP but still runs comparably fast as XP is a feat. If you are really that concerned about performance, why don't I see you using some type of DOS port? Or linux at the command line? All I am saying is that
    • by Dunbal (464142)

      Did anyone ever really consider the fact that it is a newer OS that is doing MORE than XP?

      Why yes, I did consider that perhaps this OS is constantly informing Microsoft of everything I do, tracking my credit card purchases and reading my email - but are you telling me it's true?

      Seriously, computers are so fast nowadays that any OS that can noticeably slow down your computer due to overhead is probably extremely badly written, or doing all sorts of nasty things (like "trusted computing", or encrypting DRM re

    • by Spatial (1235392)

      If they drop features to make it faster than XP then everyone will bitch that it doesn't have those features.

      People would in general, but I doubt they would here on Slashdot. A lean no-bullshit version of Windows that had much lower requirements would probably go down pretty well, doubly so if it were a lot cheaper.

  • by Anonymous Coward on Tuesday January 27, 2009 @11:55AM (#26623245)

    Actually, after looking at the benchmarks this is what I came away thinking about Vista, Win7, XP overall.

    If you want DX10 performance, Vista.
    If you want DX9 performance on iCore7, Win7.
    If you want DX9 and have a midrange system, most do, WinXP.

    The article's take on the results can be summarized in two words, "mixed bag."
    Ironically, Slashdot comes away with a bright and sunny view on things.
    Their analysis as usual does not coincide with the reality presented by the results.

    Of course, it all depends on what games you most prefer to play, for example Far Cry plays poorly in XP in all cases versus Vista/Win7.

    I find it interesting they have no benchmarks in DX10 for Fallout 3, CoD: Waw, and several more. I looked into it just now and these particular titles lack DX10 support.

    What this all means is, if you haven't upgraded to an iCore7 and most interesting games still use DX9, stick with XP. If you only play DX10 games, stay with Vista regardless what architecture you're on. Win7 fails at DX10, except in FarCry where it only does one or two fps better than Vista.

    There you go, an honest analysis of the results.

No user-servicable parts inside. Refer to qualified service personnel.

Working...