|
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <4885f9e5$1@news.povray.org>, mra### [at] hotmailcom says...
> scott wrote:
> >> Don't tell me that's the minimum requirements..
> >
> > No no no, he's only using 4% of the RAM there!
> >
>
> Oh, so ... only about 40GB of RAM, then ...
>
No, that's 40GB in the machine, the other 80GB is on the HDD. lol
Seriously, someone was doing a poorly done, but interesting comparison
on Windows vs. Linux memory usage. Windows simply crashed at a certain
point on his when swap partition was off, while Linux didn't even
notice, since it never, under the configuration it used on installation,
used it anyway. Windows of course wouldn't boot at all in 512MB, but
Linux simply refused to run the desktop. And.. memory usage of tasks...
Windows swallowed in average about 30+% more (per task), for every thing
you did with it, and in the case of IE 7, refused to give any of it back
until you closed the last window. And that doesn't even mention that
fact that, for most tasks, Linux never went above 1-2% processor
utilization, where hard tasks could take 24%, like playing video, and on
a machine without a dedicated video system for handling stuff, while
just opening the application would spike Windows at 70%+.
Basically, windows seems to be so tuned (and not possible to untune) to
swap shit out when not explicitly needed that half your processor
activity and most of your system memory is constantly being dumped back
and forth from the HDD, and that is without asking, "Why the hell does
it take 20% more to do the same thing anyway, never mind things like IE
7 never returning any of it when no longer needed? (Should be noted,
this is also the same stupid issue we ran into with Active Script. It
failed to "release" memory immediately, so every time you loaded, ran,
then dropped a small fragment of code, while there was still a main
script running in the client application, it would eat a bit of your
available ram. Do that 2,000-3,000 times in a row and your client
crashed do to sudden lack of memory, 99% of which shouldn't have even
still been "in use".)
And we wonder why it is still "unstable"? lol
--
void main () {
if version = "Vista" {
call slow_by_half();
call DRM_everything();
}
call functional_code();
}
else
call crash_windows();
}
<A HREF='http://www.daz3d.com/index.php?refid=16130551'>Get 3D Models,
3D Content, and 3D Software at DAZ3D!</A>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gail Shaw wrote:
> Nice. Wonder where I can get one of those... <g>
Why Gail, I thought you *already* had lots of big servers to play with? ;-)
[Although possibly not with a whole 2TB of RAM...]
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 22-Jul-08 22:42, Gail Shaw wrote:
> "Mike Raiford" <mra### [at] hotmailcom> wrote in message
> news:4885e8a7$1@news.povray.org...
>> Whatever machine that was running on is a BEAST.
>
> I'm glad the server team's testing out on the really, really big servers.
> The last thing you want in a server OS is a subtle bug that only manifests
> on servers with lots of CPUs/lots of memory
is that sarcasm?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"andrel" <a_l### [at] hotmailcom> wrote in message
news:488### [at] hotmailcom...
> On 22-Jul-08 22:42, Gail Shaw wrote:
> > "Mike Raiford" <mra### [at] hotmailcom> wrote in message
> > news:4885e8a7$1@news.povray.org...
>
> >> Whatever machine that was running on is a BEAST.
> >
> > I'm glad the server team's testing out on the really, really big
servers.
> > The last thing you want in a server OS is a subtle bug that only
manifests
> > on servers with lots of CPUs/lots of memory
>
> is that sarcasm?
No.
SQL Server 2005 did have a 'bug' that only manifested on servers with more
than 20 GB memory. Don't want the same to happen with the OS.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Orchid XP v8" <voi### [at] devnull> wrote in message
news:48864b76$1@news.povray.org...
> Gail Shaw wrote:
>
> > Nice. Wonder where I can get one of those... <g>
>
> Why Gail, I thought you *already* had lots of big servers to play with?
;-)
I do, but that puts my 16 proc 64GB box to shame.
Anyay, I only have access to the big servers til friday. After that the
biggest machine I'll be able to test on is my new (still in construction)
quad code, 4GB desktop.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 22-Jul-08 23:28, Gail Shaw wrote:
> "andrel" <a_l### [at] hotmailcom> wrote in message
> news:488### [at] hotmailcom...
>> On 22-Jul-08 22:42, Gail Shaw wrote:
>>> "Mike Raiford" <mra### [at] hotmailcom> wrote in message
>>> news:4885e8a7$1@news.povray.org...
>>>> Whatever machine that was running on is a BEAST.
>>> I'm glad the server team's testing out on the really, really big
> servers.
>>> The last thing you want in a server OS is a subtle bug that only
> manifests
>>> on servers with lots of CPUs/lots of memory
>> is that sarcasm?
>
> No.
>
> SQL Server 2005 did have a 'bug' that only manifested on servers with more
> than 20 GB memory. Don't want the same to happen with the OS.
>
Ah, ok, didn't know that. I did know that quite a lot of the things MS
make have trouble running on an average machine when the software comes
on the market. But that was not what you were referring to.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
andrel wrote:
> Ah, ok, didn't know that. I did know that quite a lot of the things MS
> make have trouble running on an average machine when the software comes
> on the market. But that was not what you were referring to.
>
That's due to design, though.
You can spend 12 months optimizing something to run on current hardware,
or wait 18 months and the hardware will have caught up to the point that
noone cares anymore.
...Chambers
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> a machine without a dedicated video system for handling stuff, while
> just opening the application would spike Windows at 70%+.
The spikes are pretty meaningless, because it depends on what setting you
have the update speed on in task manager. And I suspect the "update speed"
is different on Linux.
Anyway, I want my OS to use 100% CPU when I instruct it to do something
immediately, that's why I spent good money on a fast processor.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Patrick Elliott" <sel### [at] rraznet> wrote in message
>until you closed the last window. And that doesn't even mention that
>fact that, for most tasks, Linux never went above 1-2% processor
>utilization, where hard tasks could take 24%, like playing video, and on
>a machine without a dedicated video system for handling stuff, while
>just opening the application would spike Windows at 70%+.
Are you saying Linux wastes 98 - 99% of the CPU? Typical 1-2% CPU
utilization and a cap at 24% tells me that that particular system or Linux
has a severe bottleneck somewhere else.
It of course won't happen, but an ideally balanced system should show 100%
utilization in all active subsystems when performing non-interactive tasks
like launching an application, for instance. Low CPU usage can be a symptom
of poor disk caching.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>> Why Gail, I thought you *already* had lots of big servers to play with?
> ;-)
>
> I do, but that puts my 16 proc 64GB box to shame.
[Invisible wants to play with Gail's box big anyway...]
> Anyay, I only have access to the big servers til friday. After that the
> biggest machine I'll be able to test on is my new (still in construction)
> quad code, 4GB desktop.
Aww... If it makes you feel any better, *my* desktop box has 256 MB RAM
and only one core. (AthlonXP 1700 I think...)
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |