Just curious..

What speed increase could I expect by switching from a GF1 with 32MB DDR
RAM to a "plain" FX5700 with 128?

And would the graphics detail improve at the same resolution in current =
games?

TIA.

Roger.

Sort of like going from a go-cart to a formula one racing cart.
Or like trading a tricycle in for a ten speed. Do you see where I am going
with this?

"Roger" wrote in message
...
What speed increase could I expect by switching from a GF1 with 32MB DDR
RAM to a "plain" FX5700 with 128?

And would the graphics detail improve at the same resolution in current
games?

TIA.

Roger.

I thought I'd get an answer like this. But what I mean is like.. What =
speed increase
can I realisticly expect in games? How many times the frame rate if you =
will..

I compared relevant systems in the MadOnion database and it showed only =
50%
increase in points or so.. I am comparing two different versions, with =
different
level of detail, so that should mean alot, but how much exactly? If the =
2003 version
is, say twise as complex as 2001 then the FX5600 is only 4 times faster =
than GF1?
I was expecting a much bigger jump I guess. Please correct me if i'm =
wrong.

Roger.

"Dan" wrote in message =
news:6vwzb.550819$6C4.499050@pd7tw1no...
Sort of like going from a go-cart to a formula one racing cart.
Or like trading a tricycle in for a ten speed. Do you see where I am =
going
with this?
=20
"Roger" wrote in message
...
What speed increase could I expect by switching from a GF1 with 32MB =
DDR
RAM to a "plain" FX5700 with 128?
=20
And would the graphics detail improve at the same resolution in =
current
games?
=20
TIA.
=20
Roger.
=20

Depends on what you do, if you machines largest task is email and word processing you won't notice a whole lot.

--


----------------------------------------------------------------
"Roger" wrote in message ...
What speed increase could I expect by switching from a GF1 with 32MB DDR
RAM to a "plain" FX5700 with 128?

And would the graphics detail improve at the same resolution in current games?

TIA.

Roger.

"Roger" wrote in message
...

I compared relevant systems in the MadOnion database and it showed only 50%
increase in points or so.. I am comparing two different versions, with
different
level of detail, so that should mean alot, but how much exactly? If the
2003 version
is, say twise as complex as 2001 then the FX5600 is only 4 times faster
than GF1?
I was expecting a much bigger jump I guess. Please correct me if i'm wrong.


Think of it this way... why are you looking to replace your GF1? I would
assume because "newer" games are dogging it. This is because older cards
are limited by their lower clock speeds and DX compatibility (In your case
DX7) as newer games use, say DX 8 or 9 coding. Your older GF1 doesnt
support it at the video card level but renders the image using the CPU which
then relays it to the card resulting in slower performace. This is why
3dMark2001 skips the 4th game test and the shader tests.... your GF1 doesnt
support it at the card level. When I first went from a DX7 card to a DX8.1
card for example I was like "wow". I didnt realize how much detail I had
been missing with my DX8 games. Currently have a FX5600 Ultra and couldnt
be happier for what I paid for it. You mention "the FX5600 is only 4 times
faster than GF1?" that is a 400% increase over 4 years with much better
detail.... If only my pay went up at that rate.

"Roger" wrote in message
...
What speed increase could I expect by switching from a GF1 with 32MB DDR
RAM to a "plain" FX5700 with 128?

That depends on whether you have a CPU of the same era as your GF1, if that
is the case then don't expect much better performance at all. The only
difference is you'll be able to run at higher res with FSAA and things like
that on, but your CPU will be a fat bottleneck.

Roger wrote:

What speed increase could I expect by switching from a GF1 with 32MB DDR
RAM to a "plain" FX5700 with 128?

And would the graphics detail improve at the same resolution in current games?

TIA.

Roger.


With a good mid range to high end processor, you'll see a screaming
increase in performance. The old GF DDR was greatly hindered by it's
fillrate, and since then it's gone up to insane amounts. Nowadays the
GPU is mostly hindered by the total system throughput (i.e. the sum
total of your FSB, RAM latency, hard disk and chipset bandwidth) and not
the inherent limits of the card.

Also, graphics detail outside of anisotropic filtering and FSAA in
regular games won't improve compared to an older generation card with
the same current drivers very much (52.16, etc). For example, a GF2U
with the 52.16's looks the same in Unreal Tournament as a brand new GFX
5950U without FSAA and AF on.

The key difference has been in the ability to use better filtering and
much better FSAA methods than supersampling at very reasonable
performance hits instead of dropping your framerate in half. On current
FX cards, it's not unreasonable to run something at 1600x1200 in 32-bit
color with 4x antialiasing and expect good performance.

"Roger" wrote:

What speed increase could I expect by switching from a GF1 with 32MB DDR
RAM to a "plain" FX5700 with 128?

And would the graphics detail improve at the same resolution in current games?

I have a similar question. Is there a site that compares basic raw
performance of cards? I am mainly interested in the basic OpenGL
performance of gaming cards. That is, I don't care about latest eye
candy, AA and AF info would be nice but I still need an apples to
apples comparison at same resolution (which, btw, I am limited to
1280x1024 by the monitor anyway). I guess what I really need to know
is the plain/textured tri count per card per resolution per aa
setting.. etc that yield acceptable (say 20fps) performance. Any such
sites? At the moment, it's very difficult to assess price/performance
ratios and the GPU manufacturer specs (ie fillrates and such) are
pretty unreliable.

"Helion" wrote in message =
...

Think of it this way... why are you looking to replace your GF1? I =
would
assume because "newer" games are dogging it. This is because older =
cards
are limited by their lower clock speeds and DX compatibility (In your =
case
DX7) as newer games use, say DX 8 or 9 coding. Your older GF1 doesnt
support it at the video card level but renders the image using the CPU =
which
then relays it to the card resulting in slower performace. This is =
why
3dMark2001 skips the 4th game test and the shader tests.... your GF1 =
doesnt
support it at the card level. When I first went from a DX7 card to a =
DX8.1
card for example I was like "wow". I didnt realize how much detail I =
had
been missing with my DX8 games. Currently have a FX5600 Ultra and =
couldnt
be happier for what I paid for it. You mention "the FX5600 is only 4 =
times
faster than GF1?" that is a 400% increase over 4 years with much =
better
detail.... If only my pay went up at that rate.

You do have a point

Thanks alot to you and the rest for good answers! It's comforting to =
know
that it's well worth the upgrade.

The games like Max Payne 2 still run OK in 640x480 resolution on GF1 but =
I would
prefer to double it to get the details back. The FSAA on this card is =
just useless.

All in all the PC isn't state of the art anymore so I have to get it =
renewed bit by bit,
and I'll start with the gfx card. I think the CPU (PIII / 1,2 GHZ) =
should handle it
well enough till I get the opportunety to get a new mobo, ram etc..

Again, thanks.

"Darkfalz" wrote in message =
...
"Roger" wrote in message
...
What speed increase could I expect by switching from a GF1 with 32MB =
DDR
RAM to a "plain" FX5700 with 128?
=20
That depends on whether you have a CPU of the same era as your GF1, if =
that
is the case then don't expect much better performance at all. The only
difference is you'll be able to run at higher res with FSAA and things =
like
that on, but your CPU will be a fat bottleneck.

How will a PIII / 1,2GHZ handle it? Or do I need something closer to the
double of that.