PC 4GB RAM limit

  • Thread starter Thread starter Tim Anderson
  • Start date Start date
Mxsmanic said:
I'd prefer that everything be instantaneous. I've yet to see a computer
that can manage that, however, no matter how powerful. All it has to do
is manage to complete all operations in 1/60 of a second or so, which
you'd expect it to be able to do at 3+ billion operations per second;
but thanks to bloat it still cannot.
Stick 4GB of RAM in, use 1GB for the system and the other 3GB set up as
a RAM drive to which the OS and everything else runs off.
 
Mxsmanic said:
It's exceedingly rare that I hear someone say he doesn't want a GPS
receiver, and yet most people don't have one and don't want one.
Thats because they've never used one.
A lot of people tell me explicitly that they don't want a PC.

Thats because they probably use one at work and are sick of them.
PCs are expensive because they are bloated; and they have to be bloated
because software continues to swallow every improvement in hardware,
such that the only way to get reasonable response time with
currently-available software is to buy a fast, new PC.

Or use an older PC with an older OS...
 
kony said:
I suspected as much too, but when he started going on about
software bloat and cancelling performance issues, I suppose
I had simply assumed at least for a moment that the more
common Windows enhancements were disabled.

Yeah, that was my first guess too but if it's actually 'slow' then he's
either got a system problem or it's the settings.

Often that may be true, but I'd been observing it while
testing specifically for dropped frames in video, as I'd
been trying to determine how old a system could manage to
run DVD resolution divx at framerates higher than 30 FPS....
not DVDs but that's beside the point, which is that I knew
frame by frame what was there in fast action scenes and
could note instances were flashes of light weren't showing
up... then went to investigate and scrutinized it more. If
I hadn't been looking for that particular event it might
have gone unnoticed.

Oh I have no doubt you noticed it. I just meant the "it is above 1/40th"
conclusion (which you may not have meant as a 'limit' like it sounds).

If you were concluding that was the limit because you almost missed it
unless 'paying attention', people can detect an electronic flash and it's
extremely short, as short as 25 uSec.

Things get confusing real quick when talking about perception because it
depends on content and the visual cues.

I did some experiments a long time ago during the 'subliminal suggestion'
myth era and just one frame (1 of 30, say, once a minute) modified produced
noticeable 'something' with people who had no idea there was even an
experiment going on. I say 'something' because they couldn't tell what it
was but would ask "is there something wrong with the TV?"

Btw, what did you come up with as to how 'slow' a computer could handle it?
 
kony said:
I suspected as much too, but when he started going on about
software bloat and cancelling performance issues, I suppose
I had simply assumed at least for a moment that the more
common Windows enhancements were disabled.

You know, since he essentially dismissed the 100,000 to 1 cost reduction in
disks drives as not 'counting' in the performance issue he could use a
massively parallel disk array to speed things up since he'd have, by that
measure, up to 100,000 drives to play with, to hit the same price/bit mark.

Hardware interfaces for 100,000 drives might be a bit of a problem but then
1950's computers weren't $1,000 bucks either so a little extra electronics
cost wouldn't be out of line.
 
I did some experiments a long time ago during the 'subliminal suggestion'
myth era and just one frame (1 of 30, say, once a minute) modified produced
noticeable 'something' with people who had no idea there was even an
experiment going on. I say 'something' because they couldn't tell what it
was but would ask "is there something wrong with the TV?"

Ha! I knew I'd find out someday who was putting subliminal
messages in my TV set!
Btw, what did you come up with as to how 'slow' a computer could handle it?


Apathy set in before I nailed down a speed, 600MHz P3 was
too slow, 1.3GHz Tualatin Celeron handled it with a little
to spare. 1.1GHz Celeron couldn't cut it but I'm fairly
sure it was the poor Sis integrated video that was the
larger (if not only) bottleneck.
 
You know, since he essentially dismissed the 100,000 to 1 cost reduction in
disks drives as not 'counting' in the performance issue he could use a
massively parallel disk array to speed things up since he'd have, by that
measure, up to 100,000 drives to play with, to hit the same price/bit mark.

Hardware interfaces for 100,000 drives might be a bit of a problem but then
1950's computers weren't $1,000 bucks either so a little extra electronics
cost wouldn't be out of line.

He also dismissed the size of the '50s-'90 drives when
considering latency as it applied to databases... given the
low capacity of a 1990 drive, a modern beefy system could
cache the entire DB... at least that helps reads.
 
Why don't you post details on the equipment, applications, and operating
systems you work with? Perhaps that would make this a more concrete
discussion. I have no idea what you are working with. From your comments,
I'd guess it is a Pentium II 300 MHz, 64 Mbytes memory, Windows 2000, a 6
GByte hard drive that is more than 85% full, and on board video. That is
certainly the impression I get from your posts. Post the type of internet
connnection you use. How about the firewalls, antivirus protection, spyware
protection, and anti-adware protection you use.

Phil Weldon
 
David Maynard wrote in part:
Case in point is movie to TV frame drop down. If an object moves fast
enough the 2:3 frame conversion can cause some people to see it as two
objects because the eye is predicting where it should be and it isn't...
but there's 'another one just behind it' because the repeated frame left
it there when it should have moved. The primary's 'predicted' location
picks up on the next frame, so the primary object remains, but there's a
ghost following that'll be in *it's* 'predicted' location again 3 frames
later.

What is this 2:3 frame conversion for TV? As I have understood it for
years, NTSC television is 30 frames per second, 60 fields (odd lines,then
even lines, interlaced) per second. Since sound, theatrical movie projection
has been 24 frames per second. The very simplest way to convert film at 24
frame per second is to duplicate every fourth frame. That would definitely
show up as a stutter; duplicating every 4th field is slightly better. If
you watch DVDs of older British television productions ("I, Cladius", for
example) you can SEE a similar artifact resulting from the conversion of PAL
to NTSC, going from 25 frames per second to 30 frames per second. Evidently
the DVDs are made from the conversion used for the original NTSC telecasts,
when PAL/NTSC conversion used boxes that cost over $500,000 US at the time.
Thanks to the continuing development of cheaper and more capable digital
components, that kind of artifact is no longer visible. Presently the
time-wise window for the conversion can stretch over multiple frames, with
interpolation providing to-the-eye seamless image flow. Soon to come is a
complete digital path from scene to screen, with digital content becomming
an ever larger component of the scene.

Consequently, I don't understand where 2:3 comes in.

To get back to the part of this 'thread' that will not die: the 'mxsmanic'
argument. What do you suppose is the hardware/software universe in which
'mismanic' operates? I'd really like to know specifically the equipment,
os, and applications 'mxsmanic' uses. I've got to think his views are
shaped by some traumatic experiences, hopefully in the hardware/software
sphere.

Phil Weldon
 
David said:
As a number of folks have mentioned, if you don't like the bells and
whistles then just turn them off. It's a trivial thing to do.

It still would not be instantaneous. And yet it should be.
And nothing will ever be able to do it "instantaneously." The laws of
physics don't allow it.

As I've already explained, 1/60 second is good enough.
Case in point, Sachs 3D Fish screen saver won't run at speed on an old
Matrox Millennium no matter HOW fast the processor is.

Any card that can refill the screen buffer in 16 milliseconds is good
enough. The screen saver presumably depending on some sort of hardware
functions in the card, other than simple bitmaps.
 
Conor said:
Stick 4GB of RAM in, use 1GB for the system and the other 3GB set up as
a RAM drive to which the OS and everything else runs off.

I still need to write to actual disks.
 
kony said:
On certain apps there is a pause, _before_ any closing of
the window as some apps update files upon close. Otherwise,
in general no there is no lag. I'm suspecting that you've
ran out of physical memory and so are waiting on paged data
to be read?

No, I have 1.5 GB with two thirds of it idle.
 
kony said:
In a situation where the user cannot respond before the
system is ready for further user input, the system would
then seem fast enough.

This is the case for something like typing text on a screen (usually),
but for many other operations one must wait.
 
David said:
He's probably got the windows effects enabled since XP will default to that
on a 'super powerful computer' if you "let Windows chose.".

They are all disabled. I have XP configured to look like NT 4.0.
 
David said:
You know, since he essentially dismissed the 100,000 to 1 cost reduction in
disks drives as not 'counting' in the performance issue he could use a
massively parallel disk array to speed things up since he'd have, by that
measure, up to 100,000 drives to play with, to hit the same price/bit mark.

Massively parallel systems have problems of their own.
 
Phil said:
Why don't you post details on the equipment, applications, and operating
systems you work with?

No need. It's a general problem. I've never seen any exceptions. No
machine is fast enough to escape the problems.
 
kony said:
What does justification have to do with it?

Most people won't spend money without a sufficient reason for doing so.
That's a hard one to swallow. PCs are now far cheaper than
in the past for basic entry-level gear.

They are still far more expensive than they could or should be.
Then don't use new software, or make different choices.

I don't always have any choices.
The argument only works if one buys the several leaps you've
made but that are clearly false. "Some" of the newer
software does seem bloated, but that's the end of the story,
it isn't a hardship to do simple things and $1000 is just
ridiculous, there isn't anyone that can't find a PC for 1/2
that if they bother to take a few minutes to look.

Even half that price is still expensive.
 
Conor said:
Thats because they've never used one.

Many of them are compelled to use one at work. They are even less
inclined to have a PC at home.
Thats because they probably use one at work and are sick of them.

They don't see any need for a PC at home. They don't have e-mail
accounts, and they aren't interested in the Web, and PCs aren't
justifiable for much else.
Or use an older PC with an older OS...

Only new PCs with new operating systems are sold in most stores.
 
kony said:
What's your point again?
That because "some" don't want one, this means all
technology should stand still?

No. My point is that there is a big wide world out there beyond the
geeks, and almost no part of it is interested in computers, in any form.
Again, what is the point?
Random observations about minorities is kinda wasteful.

That's why geek arguments are so weak. Most people aren't geeks.
Yep, the size is a large part of it. What if you were a
woman and had enough room in your purse for a smaller
version?

I'd use the room for something else.
Transportability and weight have everything to do
with how portable something is. Further, smaller devices
more readily find their ways into new places like cars and
perhaps classroom desks.

For what purpose?

And how do you type on such small computers? How do you read the tiny
screens?
You ought to go to stores and advise them of this wisdom, as
many stores are still concluding that they are selling them.

Some sell them, some don't. The attraction is fat margins, not volume.
 
Mxsmanic said:
No need. It's a general problem. I've never seen any exceptions. No
machine is fast enough to escape the problems.
Mine doesn't seem to suffer the window resizing lag as bad as you
mention.
 
Back
Top