Windows 7 Way Smarter With Graphics RAM

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
“The elimination of the duplicate system memory copies which ‘speed up’ certain operations introduced slightly reduced performance as the CPU now has to read data back from the video memory. An analysis of real-world application statistics showed that these operations were rare,” Chitre said. “Our observation has been that these slow-downs do not impact the end-user functionality directly and that the memory savings directly result in Windows 7 being much responsive overall. The improvements overall are definitely noticeable on memory constrained PCs with shared memory graphics.”

I think this is a great step for Microsoft to create a better and more efficient OS, but I still want see some benchmarks how this will impact performance.
 
looks like MIT grads are working on win7 and highschool jocks developed vista.
kinda looks as though microsoft knowingly thru vista to the masses with all its faults, and us (the masses) wanting the next best thing bought into this fraud. microsoft did come out with win7 very fast and yes im 1 of the millions that upgraded to vista then upgraded back to xp.
im glad that microsoft at least is goin to make up for their mistake.

think they ll give anyone with a proof of purchase for vista a discount if they send it in? lol
 
[citation][nom]SkepticalSkeptic[/nom]See, this is how they avoid "perception problems" like in Vista, have the news constantly flood you with puff pieces on what a great OS this is, and thus, rumors that it doesn't suck spread like wildfire. This is why it's a new OS rather than Vista SP2.5, even though it's based on Vista. I hope it lives up to the hype, but don't let them "market" it to you like Apple does.[/citation]
Woah, woah, woah.

Apple prays on ignorant consumers with subjective comparisons and outright lies. Microsoft is reporting on objective technical features they've added to their operating system. Ignorant consumers aren't even going to understand, much less be swayed by such reports.
 
[citation][nom]Naw-yi[/nom]and yes im 1 of the millions that upgraded to vista then upgraded back to xp.[/citation]

That's not the kind of thing you should admit on a blog where people tend to actually know stuff...
 
Surely MS off-loaded shadow video RAM ages ago, with a Hotfix to Vista (now incorporated in SP1)? See this KB article:

http://support.microsoft.com/kb/940105

I quote:

"With the introduction of DirectX 10 and Windows Display Driver Model (WDDM) in Windows Vista, it is no longer necessary for an application to maintain a copy of its resources in system memory. Instead, the video memory manager makes sure that the content of every video memory allocation is maintained across display transitions. For compatibility reasons, Windows Vista emulates "device lost" for DirectX versions that are earlier than DirectX 10 to make sure that no application-visible API behavior changes."

It sounds to me as if MS are spinning Windows 7 by comparing it with Vista RTM rather than Vista as "mended" by the hotfix and/or SP1.

There is a danger that this and other websites have become un-moderated conduits for the manufacturers to promote their marketing spin. So I hope the issue will be looked into. And if it turns out that MS is just re-spinning an existing technology, I hope that Tom's will point it out in forceful terms.
 
Here's a thought.... how about not making bloatware that gobbles up memory like there's not tomorrow - then you wouldn't need this "advancement". The explosion in memory consumption by the Window OS over the last few generation is a large part of why this is being touted as a good thing.

While the change seems good, if everything was written to use memory more efficiently in the first place then removing this redundancy would not be a big deal. I am curious whether it is 'only benchmarks' that are potentially slowed down as the CPU now has to get the data from the GPU memory.
 
Feeding the m$ fanboys with more marketing BS by the spoon, and all start again to "salivate over a new piece of technology".
While it makes sense for shared RAM configurations (the video RAM address space was anyway directly accessible by the CPU), for dedicated video RAM it'll mean just more overhead - but this one is easily masked by the huge DRM overhead built-in in vi$hta RTM/SP1/SP2/SP2+(aka $even).
While the "success" story of windblow$ memory management is well known, now m$ is starting to bloat over in video RAM, truly dedicated to their strategy of clogging all available resources (wasn't that a virus "feature"?).
 
I really thought they were doing that before... It's not as if it's a new idea!

For those of you not in the know, X11 is the protocol used to draw GUIs under UNIX OSes (apart from Mac OS X). Its 'legacy' acceleration method, called XAA, dates back to the 1990's. It has been mostly replaced with EXA (which has been extended to make use of a 3D engine to accelerate 2D operations). Intel has taken EXA and removed the part that requires CPU copies and created UXA based off EXA.

By reading Keith Packard's blog, it seems that the UNIX community (at least under BSD and Linux, but probably Solaris and others too) has been saving RAM like that for a looong while (see his description of how EXA manages pixmap writes) and has KNOWN the issue with such a system: when the CPU has to read back from VRAM, it is horribly expensive.

So, this system is viable only if:
- you have a VERY fast PCIe card, with a very fast bus
- you have TRUCKLOADS of VRAM, so as to never need to move pixmaps back and forth
- you have dynamically allocated RAM used as VRAM
And it depends on:
- having a very efficient RAM/VRAM allocator, that can reduce movements between VRAM and RAM as efficiently as possible.

Look at the graph above: opening a window eats 40 Mb. Open a dozen: there, 500 Mb of VRAM allocated.

But you have only 256 Mb of VRAM...

Well, it means that the OS has to perform pixmap swaps. These can happen on something as stupid as a glyph (a text element). Imagine that the VRAM allocator hasn't freed enough RAM to hold a glyph set (and a glyph isn't a single byte: it's a full matrix of dots, of which we only see a filtered subsample on screen due to kerning, hinting and such shit, and it can't be stored in VRAM as a path), and you type fast.

Time travel back to the 1990's, you'll see the glyphs draw themselves on screen as you type.
 
@ horseouttabarn and ossie

Before opening your mouth and removing all doubt that you have no idea wth you are talking about, why don't you actually try the new OS instead of regurgitating FUD that's 2 years old. Might actually help you make a valid point now and then. Ahh but who am I kidding. Folks like that are impervious to logic. SHould've known better and not wasted my time.
 
[citation][nom]tfm[/nom]There is a danger that this and other websites have become un-moderated conduits for the manufacturers to promote their marketing spin. So I hope the issue will be looked into. And if it turns out that MS is just re-spinning an existing technology, I hope that Tom's will point it out in forceful terms.[/citation]

Totally agree.
 
I like what I hear. Then again, everything i heard about Vista was positive, and I went out and dropped money for two ultimate editions that I never use. I am going to look at all these things as positive, but will likely spend some time playing with windows 7 on someone elses computer before I buy into the hype this time. So far Microsoft seems to be on the ball.
 
"This change positively impacts real-world usability, but benchmarks may show a degradation in performance since the CPU has to fetch data from video RAM."

This snippet of the article implies that fetching from graphics RAM is slower than from system RAM. While I see the advantage this modification provides for power users who like to keep 8 apps running concurrently, won't this mean a performance hit for gamers?
 
@mitch074: You do realize that most people likely to use that many windows will most likely be a power user or enthusiast of sorts, right? That being said, the user will most likely have a system that hits all the conditions that you listed. So.. moot point. Not to mention you ignore the number of page faults introduced by duplicate data and that's going to a physical disk. Why don't you factor that into your "big overhead"?

@Fadamor: If you're a gamer I doubt you're running something in windowed mode, but rather fullscreen so the rendering method and handling of memory would change. This is dealing strictly with the 2D, desktop graphics stack. I doubt that gamers would see a negative impact.
 
[citation][nom]Pei-chen[/nom]Play 1080p file with XP and Vista x64 and you'll see the difference.[/citation]

This is dependant on the codec, not the OS.
 
@TindyTim - "Apple prays on ignorant consumers"

No, Apple prays that their customers are ignorant so that they can prey on them :)
 
[citation][nom]deck[/nom]This is dependant on the codec, not the OS.[/citation]
It's dependent on the OS and Hardware as well. UVDs are prominent nowadays, meaning there is no longer a need for software codecs.
 
Sounds like maybe this needs to be configurable. If you have memory to spare why not use it? If not configurable then it could adjust on the fly as memory usage hits some percentage.

Anyway, it's cool that MS is showing us a few little details about how things are changing with Win7.
 
[citation][nom]smlong[/nom]I can run multiple 1080P videos on my Q9550 system running either Vista x64 or Win7 x64. My CPU usage running (2) clips is well under 50% utilization.[/citation]
So? What does that have to do with my statement?
 
[citation][nom]smlong[/nom]Well, I guess you have to follow the entire thread to which your message was attached.[/citation]
I was, you just seemed to blurt out some random thing. Saying your CPU usage is only at a certain level doesn't say anything without any background. You mentioned 2 64-bit OSes, and didn't mention you GPU (which could easily have UVDs).
 
Status
Not open for further replies.