![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Anandtech Vista Performance Guide
- Thread starter Archer75
- Start date
I've kinda relied on Anand's reviews and articles (meaning the website) over the years because they make sense and don't get buried under a lot of personal opinions and bias. But one thing in that review/guide I'll quote here:
"Benchmarking in Vista is an extremely difficult task, thanks to features like SuperFetch designed to improve performance at the sacrifice of repeatability. We're in love with SuperFetch as a feature, but it does make our jobs a bit more difficult as we have to be very mindful of what has been cached into main memory and how that impacts the performance of our benchmarks."
Now, call me silly and slap with with a trout (old IRC joke) but, considering the fact that Vista is designed to use SuperFetch and ReadyBoost and all these other newfangled technologies to make it work more efficiently and perform better than XP and previous versions of Windows, why go the unneeded step of making it sound as though those and other new features that make Vista better sound almost like a handicap? Does that make sense?
I read that paragraph as saying:
"Well, we wanted to get reliable performance data of XP vs Vista vs Vista x64, but since Vista and Vista x64 seem to have an unfair advantage of actually using system RAM to make itself much faster and more efficient, we basically had to kneecap Vista and Vista x64 to make them work more like XP and therefore keep the benchmarks more realistically balanced."
I mean, I just lost a lot of respect for Anand's site from just those two sentences I quoted above in italics. They're saying they almost need to handicap Vista and Vista x64 so it looks like a fair comparison because Vista and Vista x64 will use available system RAM to dramatically improve performance in day to day use and applications over XP's performance in the same day to day use and the same applications.
I'm pretty disappointed in that review now, just thought I'd point out why. Damned disappointed, actually. It's not a fair comparison in my eyes.
While the test results still show XP with a slight advantage overall on average, meaning better in some places by a noticeable margin, slower in others, and in most it's just barely faster, the fact that they'd come out and say those two lines above just blows my mind.
It's like saying "Little Johnny here can run, and Little Tommy can't because he's in a wheelchair, so to make it fair we handicapped Johnny by weighing him down with diver's boots and a lead belt."
Silly analogy, but that's how the image in my imagination comes to mind when I think about just those two sentences.
"Benchmarking in Vista is an extremely difficult task, thanks to features like SuperFetch designed to improve performance at the sacrifice of repeatability. We're in love with SuperFetch as a feature, but it does make our jobs a bit more difficult as we have to be very mindful of what has been cached into main memory and how that impacts the performance of our benchmarks."
Now, call me silly and slap with with a trout (old IRC joke) but, considering the fact that Vista is designed to use SuperFetch and ReadyBoost and all these other newfangled technologies to make it work more efficiently and perform better than XP and previous versions of Windows, why go the unneeded step of making it sound as though those and other new features that make Vista better sound almost like a handicap? Does that make sense?
I read that paragraph as saying:
"Well, we wanted to get reliable performance data of XP vs Vista vs Vista x64, but since Vista and Vista x64 seem to have an unfair advantage of actually using system RAM to make itself much faster and more efficient, we basically had to kneecap Vista and Vista x64 to make them work more like XP and therefore keep the benchmarks more realistically balanced."
I mean, I just lost a lot of respect for Anand's site from just those two sentences I quoted above in italics. They're saying they almost need to handicap Vista and Vista x64 so it looks like a fair comparison because Vista and Vista x64 will use available system RAM to dramatically improve performance in day to day use and applications over XP's performance in the same day to day use and the same applications.
I'm pretty disappointed in that review now, just thought I'd point out why. Damned disappointed, actually. It's not a fair comparison in my eyes.
While the test results still show XP with a slight advantage overall on average, meaning better in some places by a noticeable margin, slower in others, and in most it's just barely faster, the fact that they'd come out and say those two lines above just blows my mind.
It's like saying "Little Johnny here can run, and Little Tommy can't because he's in a wheelchair, so to make it fair we handicapped Johnny by weighing him down with diver's boots and a lead belt."
Silly analogy, but that's how the image in my imagination comes to mind when I think about just those two sentences.
HotShotMedic
Gawd
- Joined
- Aug 3, 2006
- Messages
- 651
good article. for anyone thinking about upgrading to vista (or those who did and don't see what all the fuss is about) i suggest reading it.
As I Lay Dying
Gawd
- Joined
- Aug 8, 2005
- Messages
- 975
http://www.tomshardware.com/2007/01/29/xp-vs-vista/page3.html
well here is a tom review and it seems they trained and had superfetch enabled, and vista got shit canned in some benchmarks.
Again I do not want to jump to a conclusion but sufficed to say is it fair to run apps that are not really designed for vista?
well here is a tom review and it seems they trained and had superfetch enabled, and vista got shit canned in some benchmarks.
Again I do not want to jump to a conclusion but sufficed to say is it fair to run apps that are not really designed for vista?
What it sounds like to me is that AT is saying that Superfetch will make it more difficult to get benchmarks to be consistent, just running in Vista alone. Forget about the comparisons to XP.
You may get varying results depending on what's in the cache. A game may run faster after running several benches than it was on the first one, etc.
True, true, but that's just a good thing all around in my perspective. I doubt you can adequately test how well SuperFetch and ReadyBoost might assist a machine in any consistent way, unfortunately, so... benchmarks be damned. If it's fast, it's fast, 'nuff said.
As for the SPECviewperf stuff, OpenGL is effectively crippled in Vista in many ways. If professionals doing high end OpenGL graphics are serious about their work, they won't be using Vista anytime soon, unless Microsoft caves in and comes out with some modification or update to Vista, or allows the hardware vendors like ATI, Nvidia, or even 3DLabs to find some way to work around the negatives in Vista regarding OpenGL support.
We'll see what happens...
As for the SPECviewperf stuff, OpenGL is effectively crippled in Vista in many ways. If professionals doing high end OpenGL graphics are serious about their work, they won't be using Vista anytime soon, unless Microsoft caves in and comes out with some modification or update to Vista, or allows the hardware vendors like ATI, Nvidia, or even 3DLabs to find some way to work around the negatives in Vista regarding OpenGL support.
We'll see what happens...