Windows 7 Memory Usage

For those of us who saw a recent Computer World article claiming
that "Most Windows 7 PCs max out memory", I see that Win7News has
just pointed to a couple articles discrediting the "finding":

<http://www.osnews.com/story/22896/Windows_7_Memory_Usage_FUD_Explained>
<http://blogs.zdnet.com/BTL/?p=31024>

Computer World has left the story on line, with a disclaimer:
<http://www.computerworld.com/s/article/9158258/Most_Windows_7_PCs_max_out_memory>

--
Dave

Dave Rudisill <> wrote in
news::

> For those of us who saw a recent Computer World article claiming
> that "Most Windows 7 PCs max out memory", I see that Win7News has
> just pointed to a couple articles discrediting the "finding":
>
> <http://www.osnews.com/story/22896/Wi..._FUD_Explained
> > <http://blogs.zdnet.com/BTL/?p=31024>
>
> Computer World has left the story on line, with a disclaimer:
> <http://www.computerworld.com/s/artic...ows_7_PCs_max_
> out_memory>
>


Huh. Looks like they got serious egg on their faces. I'm looking at my
machine, Win7 64 bit, 4GB ram.

Used: 2250
Cached: 1830
Available: 1836
Free: 41

They look at free, not available, and came to some bad conclusions.

I just loaded an ~80MB file into memory, and this is how it changed.

Used: 2341 (+91)
Cached: 1780 (-50)
Available: 1755 (-81)
Free: 4 (-37)

(the above numbers fluctuate a bit, but not by much. As I typed this,
available went up to 1763, free went up to 16)

Used memory went up 91MB, pretty much expected. Cached went down 50, free
went down 37. It looks like it took everything that was "free", and then
started taking cached to make up the difference. All of this was
lightning fast. If I load 500MB into memory, it takes it out of cached -
lightning fast. If I unload it, it goes into free. And then free slowly
drops and cached slowly increases until it reaches some sort of
equilibrium. Very interesting.

So, yeah, they seriously screwed up by crying the sky is falling before
taking the time to do their homework.

"Zootal" <> wrote in message
news:Xns9D2A87CBBD3C4nospamspamzootalnosp@216.196.97.131...
> Dave Rudisill <> wrote in
> news::
>
>> For those of us who saw a recent Computer World article claiming
>> that "Most Windows 7 PCs max out memory", I see that Win7News has
>> just pointed to a couple articles discrediting the "finding":
>>
>> <http://www.osnews.com/story/22896/Wi..._FUD_Explained
>> > <http://blogs.zdnet.com/BTL/?p=31024>
>>
>> Computer World has left the story on line, with a disclaimer:
>> <http://www.computerworld.com/s/artic...ows_7_PCs_max_
>> out_memory>
>>
>
>
> Huh. Looks like they got serious egg on their faces.

Especially since SuperFetch has been around since Vista.


--

-smithdoerr

On 2010-02-26 00:59:54 +0000, smithdoerr said:

>
> Especially since SuperFetch has been around since Vista.

Indeed. Superfetch is so good that they introduced a similar feature
for Linux. Whats the point of having 4GB in a machine if for 75% of the
time, 75% of it sits there empty?
--
Conor

i'm not prejudiced, I hate everyone equally.

On 2010-02-26 00:59:54 +0000, smithdoerr said:

> "Zootal" <> wrote in message
> news:Xns9D2A87CBBD3C4nospamspamzootalnosp@216.196.97.131...
>> Dave Rudisill <> wrote in
>> news::
>>
>>> For those of us who saw a recent Computer World article claiming
>>> that "Most Windows 7 PCs max out memory", I see that Win7News has
>>> just pointed to a couple articles discrediting the "finding":
>>>
>>> <http://www.osnews.com/story/22896/Wi..._FUD_Explained
>>>> <http://blogs.zdnet.com/BTL/?p=31024>
>>>
>>> Computer World has left the story on line, with a disclaimer:
>>> <http://www.computerworld.com/s/artic...ows_7_PCs_max_
>>> out_memory>
>>>
>>
>>
>> Huh. Looks like they got serious egg on their faces.
>
> Especially since SuperFetch has been around since Vista.


--
Conor

i'm not prejudiced, I hate everyone equally.

Conor <> wrote in news:hm8hnl$otv$-
september.org:

> On 2010-02-26 00:59:54 +0000, smithdoerr said:
>
>>
>> Especially since SuperFetch has been around since Vista.
>
> Indeed. Superfetch is so good that they introduced a similar feature
> for Linux. Whats the point of having 4GB in a machine if for 75% of the
> time, 75% of it sits there empty?

Is this in the kernel? Or is it and addon? What distro, what version?

Dave Rudisill wrote:
> For those of us who saw a recent Computer World article claiming
> that "Most Windows 7 PCs max out memory", I see that Win7News has
> just pointed to a couple articles discrediting the "finding":
>
> <http://www.osnews.com/story/22896/Windows_7_Memory_Usage_FUD_Explained>
> <http://blogs.zdnet.com/BTL/?p=31024>
>
> Computer World has left the story on line, with a disclaimer:
> <http://www.computerworld.com/s/article/9158258/Most_Windows_7_PCs_max_out_memory>
>


If the software is the type of software that I think it is, then these
sorts of software have been around since the beginning of Windows on
DOS. It's simply something that takes unprotected pages from memory and
commits it to disk, thus miraculously freeing up the memory. Then of
course through normal computer utilization, that freed memory gets eaten
up again by the programs that are still running and still require that
memory. It doesn't really buy you anything, it's just a placebo.

They used to have software like this in the DOS days that would optimize
the memory of DOS apps and background programs. DOS was severely limited
in the amount of memory it could use, and also in DOS the whole program
would have to be loaded into memory right away, and not partially on
demand like they do in Windows. So when Windows came around, people were
still used to being stingy with memory and many used to personally
monitor it like a hawk, so utilities came out that purported to do the
same thing for Windows memory that they did for DOS memory. It would
simply commit unsaved memory to disk and close it down, only to have the
memory get reutilized after a couple of hours. I even made one such
utility for myself as a programming exercise. It was pretty easy to do,
the function was built right into Windows, all you had to do was create
a fancy interface to put around the function.

Yousuf Khan