How much of a speed increase?

'Mr. E Solved!'
| The value of increased FPS has zero to do with "seeing" 1/1000th (or any
| other reasonable fraction) of second state changes in the monitor.
|
| Phosphors in displays have a medium-fast decay time that negate hyper
| fast visual changes. If you want to switch a pixel from purple to green
| and back in 1/500th of a second repeatedly, you need a different
technology.
|
| High FPS desirability has everything to do with being able to accurately
| reproduce and synchronize your client gaming environment with the Master
| Server environment. Add to that burdensome task maintaining that steady,
| bright and vivid image on your display. Refreshing those phosphors as
| intended. Then add that task the I/O of transferring those state changes
| in and out of your PC.
_____

You evidently did not read my post "At most, an FPS rating may indicate the
EXCESS CAPACITY of a CPU/Memory System/Graphics System) for displaying a
scripted section of a game. An EXCESS CAPACITY that may or may
not ever be called on" .and you fail to understand that no matter HOW fast a
pixel can change, at a certain point the switch from one frame to the next
in a motion sequence is perceived as continuous motion. My
recommendations -

read a post before replying

quote a post when you reply (might give you a second chance to catch the
meaning)

understand a concept before trying to use it or explain it

pick a more apt sig.

Phil Weldon

"Mr.E Solved!" wrote in message
. ..
| Phil Weldon wrote:
|
| Well, the FPS rating is not exactly linear as a comparison benchmark.
After
| all, who can SEE 1000 frames per second (or 100).
|
| The value of increased FPS has zero to do with "seeing" 1/1000th (or any
| other reasonable fraction) of second state changes in the monitor.
|
| Phosphors in displays have a medium-fast decay time that negate hyper
| fast visual changes. If you want to switch a pixel from purple to green
| and back in 1/500th of a second repeatedly, you need a different
technology.
|
| High FPS desirability has everything to do with being able to accurately
| reproduce and synchronize your client gaming environment with the Master
| Server environment. Add to that burdensome task maintaining that steady,
| bright and vivid image on your display. Refreshing those phosphors as
| intended. Then add that task the I/O of transferring those state changes
| in and out of your PC.
|
| Battlefield 2 servers are digital worlds that update 100 times a second.
| they live in a 100Hz universe. If your client PC can match 100FPS
| sustained rate and re-draw the display 100 times a second then you have
| achieved perfect game-state harmony and the most accurate (barring
| latency spikes) arena possible.
|
| If I have 100 FPS sustained and you have 50 FPS sustained, I have twice
| as many opportunities as you do and the score will reflect that. (all
| else being equal). To answer your question then: Everyone can see the
| difference when it is shown to them, but not every one can explain it.
|
| HTH.

"Rich" wrote in message
...
On Fri, 19 Oct 2007 13:28:01 -0400, "Mr.E Solved!"
wrote:

Rich wrote:


I've recently upgraded from a 7600G to 7900 GTX. A 7950 is basically
2 x 7900 GTX stuck together.

No, a 7950GT is a tweaked 7900GT. A 7950GX2 is more along the lines of
two 7900GT's SLI'ed rather than two 7900GTx's SLI'ed.
The 7950GX2 has a slightly faster core clock (and slower memory cock)
than a 7900GT, and does not approach 7900GTX speed.

Thanks for that. Was tempted on a 7950 GX2 to replace my existing
card. But looks like its not worth it then.

Rich

I own one... and I don't think I'll be bothering with a faux-SLI card like
it, or real SLI in the future. Just not worth the expense, when a lot games
don't see much benefit from SLI, due to the rendering modes they use, among
other things.

RF.

Phil Weldon wrote:

at a certain point the switch from one frame to the next
in a motion sequence is perceived as continuous motion.

Which is entirely irrelevant, this isn't cinema. This is an interactive
environment. Continuous motion is perceived at very low frame rates, far
lower then you might expect, ever hear of motion blur?

Your goal of "continuous motion" isn't even met in your cinema analogy.
If you had a rapidly rotating reflective/rebound object like a wall,
which rotates at a certain RPM, say FASTER than your FPS, you will never
be able to aim your shot off of the rotating surface, as you will never
see it where it needs to be "in continuous motion or not".

How about driving in a jeep down a road, picket fence on the drivers
side? You are shooting, oh I don't know...a machine gun. If your FPS
doesn't match the rate of change as you drive down the road, you will
never be able to shoot through the picket fence with predictable
accuracy since you will never see where the opening actually is, even if
the scene progresses smoothly or in continuous motion.

Yes, a CRT showing a static image of the Mona Lisa appears no different
to the "human eye" (as opposed to a walrus eye?) at 10Hz, 100Hz or
1000Hz depending on the half-life of the phosphors. But as soon as you
put Mona in motion, in a helicopter or in sword fighting combat, you
will find every single FPS is crucial, and the basis of all benchmarks.

I'm sorry you don't understand how it all works, but it does and that's
why we have very high refresh rates and that's why we use them.

Stop resisting deeper knowledge, it hurts us all in the end.

'Mr.E Solved!' wrote:
| Which is entirely irrelevant, this isn't cinema. This is an interactive
| environment. Continuous motion is perceived at very low frame rates, far
| lower then you might expect, ever hear of motion blur?
_____

A mind is a terrible thing to waste. With no relevant experience or
specific knowledge, you, sadly, have nothing worthwhile to say. But persist
in posting anyway. As just one example, you fail to understand the concept
of 'motion blur' (motion blur can occur even in a still picture - it is a
artifact of movement while the shutter is open. The equivalent for video is
a horizontally spread image for interlaced video and a tilted image for
non-interlaced video. For more information I suggest you read, experiment,
or ask someone who knows.

Phil Weldon


"Mr.E Solved!" wrote in message
. ..
| Phil Weldon wrote:
|
| at a certain point the switch from one frame to the next
| in a motion sequence is perceived as continuous motion.
|
| Which is entirely irrelevant, this isn't cinema. This is an interactive
| environment. Continuous motion is perceived at very low frame rates, far
| lower then you might expect, ever hear of motion blur?
|
| Your goal of "continuous motion" isn't even met in your cinema analogy.
| If you had a rapidly rotating reflective/rebound object like a wall,
| which rotates at a certain RPM, say FASTER than your FPS, you will never
| be able to aim your shot off of the rotating surface, as you will never
| see it where it needs to be "in continuous motion or not".
|
| How about driving in a jeep down a road, picket fence on the drivers
| side? You are shooting, oh I don't know...a machine gun. If your FPS
| doesn't match the rate of change as you drive down the road, you will
| never be able to shoot through the picket fence with predictable
| accuracy since you will never see where the opening actually is, even if
| the scene progresses smoothly or in continuous motion.
|
| Yes, a CRT showing a static image of the Mona Lisa appears no different
| to the "human eye" (as opposed to a walrus eye?) at 10Hz, 100Hz or
| 1000Hz depending on the half-life of the phosphors. But as soon as you
| put Mona in motion, in a helicopter or in sword fighting combat, you
| will find every single FPS is crucial, and the basis of all benchmarks.
|
| I'm sorry you don't understand how it all works, but it does and that's
| why we have very high refresh rates and that's why we use them.
|
| Stop resisting deeper knowledge, it hurts us all in the end.
|

Phil Weldon wrote:

As just one example, you fail to understand the concept
of 'motion blur' (motion blur can occur even in a still picture - it is a
artifact of movement while the shutter is open. The equivalent for video is
a horizontally spread image for interlaced video and a tilted image for
non-interlaced video. For more information I suggest you read, experiment,
or ask someone who knows.


Stop working! All you digital artists, level creators, halt! Graphic
engineers, software coders and systems analysts...hold on! Phil says to
you all that high FPS can't be seen, in fact they are wasted when they
get past a point Phil has yet to determine.

What's that? You want me to ask Phil what refresh rate is best? That
sacred ceiling that we shant go faster than, the Weldon Limit? Ok Phil,
here's your chance to school us all, since the combined brain-trust of
the game world is doing it all wrong. What is the Weldon Limit of FPS?

Hurry with a response, there are people coding "WW1 fighters" who can't
complete the through-the-propeller-firing machine gun on the Sopwith
camel. Because a rotating prop model doesn't have any need to go faster
than the Weldon Limit.

This isn't photography, the rules of rendering are different. Feel
fortunate, gentle reader, that unskilled people who think like Phil do
not work in the digital field making your games/hardware.

Phil Weldon wrote:
'Moo' wrote:
| Nice, thanks for that... might be investing in one of those!
_____

Well, the FPS rating is not exactly linear as a comparison benchmark. After
all, who can SEE 1000 frames per second (or 100). At most, an FPS rating
may indicate the EXCESS CAPACITY of a CPU/Memory System/Graphics System) for
displaying a scripted section of a game. An EXCESS CAPACITY that may or may
not ever be called on. [Not to mention that 160% does not even seem to be a
number that can be extracted from the cited page!]

Benchmarks are of limited value, ESPECIALLY when improperly interpreted.
And as long as "when I do decide to upgrade the PC completely I'm gonna have
to upgrade every component has technology has moved on drastically" then
surely the non DX10 capable 7950 GT would not be in the cards. What SLI
gives you is a choice - install one card, and, if game play is not
satisfying to you, then adding a second, identical card is an easy upgrade.

Unless you are going for bragging rights evaluation of game play is much
more useful than comparing benchmarks.

Phil Weldon


Erm, what?

I don't have an SLI board, just a standard AGP hence I will need to
upgrade everything.

I know that the 7950 doesn't do DX10, if I wanted DX10 I would upgrade
everything and get the appropriate card for that, as it stands I'm just
trying to get more time out of my current PC for as little as possible.

--

My blog - http://degeneratemoo.livejournal.com

'Mr.E Solved!' wrote, in part:

| This isn't photography, the rules of rendering are different. Feel
| fortunate, gentle reader, that unskilled people who think like Phil do
| not work in the digital field making your games/hardware.
_____

You still haven't read my first post, evidently (what do you think "may or
may not ever be called on" means?). How about this, why not explain
something that you do know about? Then judging the value of your comments
might be more fair to you. Also, your 'replies' might be easier to judge if
they were actually on point.

Phil Weldon

"Mr.E Solved!" wrote in message
. ..
| Phil Weldon wrote:
|
| As just one example, you fail to understand the concept
| of 'motion blur' (motion blur can occur even in a still picture - it is
a
| artifact of movement while the shutter is open. The equivalent for
video is
| a horizontally spread image for interlaced video and a tilted image for
| non-interlaced video. For more information I suggest you read,
experiment,
| or ask someone who knows.
|
|
| Stop working! All you digital artists, level creators, halt! Graphic
| engineers, software coders and systems analysts...hold on! Phil says to
| you all that high FPS can't be seen, in fact they are wasted when they
| get past a point Phil has yet to determine.
|
| What's that? You want me to ask Phil what refresh rate is best? That
| sacred ceiling that we shant go faster than, the Weldon Limit? Ok Phil,
| here's your chance to school us all, since the combined brain-trust of
| the game world is doing it all wrong. What is the Weldon Limit of FPS?
|
| Hurry with a response, there are people coding "WW1 fighters" who can't
| complete the through-the-propeller-firing machine gun on the Sopwith
| camel. Because a rotating prop model doesn't have any need to go faster
| than the Weldon Limit.
|
| This isn't photography, the rules of rendering are different. Feel
| fortunate, gentle reader, that unskilled people who think like Phil do
| not work in the digital field making your games/hardware.
|
|
|
|
|

'Moo' wrote:
| Erm, what?
|
| I don't have an SLI board, just a standard AGP hence I will need to
| upgrade everything.
|
| I know that the 7950 doesn't do DX10, if I wanted DX10 I would upgrade
| everything and get the appropriate card for that, as it stands I'm just
| trying to get more time out of my current PC for as little as possible.
_____

Since you will eventually be updating the entire system, it is even more
important to get the best value DX9 AGP board that will meet your gaming
requirements. If you'd share your time frame for a more complete upgrade I'm
sure you'll get a lot more suggestions. (A basic SLI nVidia motherboard, a
Core 2 Duo CPU, and 2 GBytes of DDR2 memory costs less than a 8800 GT PCI-E
video card - I've got an Intel 4300 Core 2 Duo @ 2.7 GHz, two GBytes
DDR2-1066 memory, an EVGA 680i SLI motherboard, and an EVGA 8800 GTS 320
MByte PCI-E card that gets about 10,000 on 3DMark06 with overclocking. A
reasonable upgrade path could include a faster Core 2 Duo [or Core 2 Duo
Quad] and/or a second 8800 GTS or two 8800 GT cards.

Raw frame rates don't exactly equal increased gaming capability across the
board of requirements. For the same game used in a FPS benchmark, but with
increased on screen and off screen complexity, extra FPS benchmark score on
the original scripted game benchmark with less complex will not translate
directly into increased game performance. A better basis to choose is a
description of, or experience with a particular display adapter and a
particular game.

Phil Weldon


"Moo" wrote in message
...
| Phil Weldon wrote:
| 'Moo' wrote:
| | Nice, thanks for that... might be investing in one of those!
| _____
|
| Well, the FPS rating is not exactly linear as a comparison benchmark.
After
| all, who can SEE 1000 frames per second (or 100). At most, an FPS
rating
| may indicate the EXCESS CAPACITY of a CPU/Memory System/Graphics System)
for
| displaying a scripted section of a game. An EXCESS CAPACITY that may or
may
| not ever be called on. [Not to mention that 160% does not even seem to
be a
| number that can be extracted from the cited page!]
|
| Benchmarks are of limited value, ESPECIALLY when improperly interpreted.
| And as long as "when I do decide to upgrade the PC completely I'm gonna
have
| to upgrade every component has technology has moved on drastically" then
| surely the non DX10 capable 7950 GT would not be in the cards. What SLI
| gives you is a choice - install one card, and, if game play is not
| satisfying to you, then adding a second, identical card is an easy
upgrade.
|
| Unless you are going for bragging rights evaluation of game play is much
| more useful than comparing benchmarks.
|
| Phil Weldon
|
|
| Erm, what?
|
| I don't have an SLI board, just a standard AGP hence I will need to
| upgrade everything.
|
| I know that the 7950 doesn't do DX10, if I wanted DX10 I would upgrade
| everything and get the appropriate card for that, as it stands I'm just
| trying to get more time out of my current PC for as little as possible.
|
| --
|
| My blog - http://degeneratemoo.livejournal.com

Phil Weldon wrote:
'Moo' wrote:
| Erm, what?
|
| I don't have an SLI board, just a standard AGP hence I will need to
| upgrade everything.
|
| I know that the 7950 doesn't do DX10, if I wanted DX10 I would upgrade
| everything and get the appropriate card for that, as it stands I'm just
| trying to get more time out of my current PC for as little as possible.
_____

Since you will eventually be updating the entire system, it is even more
important to get the best value DX9 AGP board that will meet your gaming
requirements. If you'd share your time frame for a more complete upgrade I'm
sure you'll get a lot more suggestions. (A basic SLI nVidia motherboard, a
Core 2 Duo CPU, and 2 GBytes of DDR2 memory costs less than a 8800 GT PCI-E
video card - I've got an Intel 4300 Core 2 Duo @ 2.7 GHz, two GBytes
DDR2-1066 memory, an EVGA 680i SLI motherboard, and an EVGA 8800 GTS 320
MByte PCI-E card that gets about 10,000 on 3DMark06 with overclocking. A
reasonable upgrade path could include a faster Core 2 Duo [or Core 2 Duo
Quad] and/or a second 8800 GTS or two 8800 GT cards.

Raw frame rates don't exactly equal increased gaming capability across the
board of requirements. For the same game used in a FPS benchmark, but with
increased on screen and off screen complexity, extra FPS benchmark score on
the original scripted game benchmark with less complex will not translate
directly into increased game performance. A better basis to choose is a
description of, or experience with a particular display adapter and a
particular game.

Phil Weldon


"Moo" wrote in message
...
| Phil Weldon wrote:
| 'Moo' wrote:
| | Nice, thanks for that... might be investing in one of those!
| _____
|
| Well, the FPS rating is not exactly linear as a comparison benchmark.
After
| all, who can SEE 1000 frames per second (or 100). At most, an FPS
rating
| may indicate the EXCESS CAPACITY of a CPU/Memory System/Graphics System)
for
| displaying a scripted section of a game. An EXCESS CAPACITY that may or
may
| not ever be called on. [Not to mention that 160% does not even seem to
be a
| number that can be extracted from the cited page!]
|
| Benchmarks are of limited value, ESPECIALLY when improperly interpreted.
| And as long as "when I do decide to upgrade the PC completely I'm gonna
have
| to upgrade every component has technology has moved on drastically" then
| surely the non DX10 capable 7950 GT would not be in the cards. What SLI
| gives you is a choice - install one card, and, if game play is not
| satisfying to you, then adding a second, identical card is an easy
upgrade.
|
| Unless you are going for bragging rights evaluation of game play is much
| more useful than comparing benchmarks.
|
| Phil Weldon
|
|

I will be putting off the upgrade for as long as possible as I simply
couldn't afford to replace everything in the PC, for £140 I can get
reasonable performance with the system I have for a while yet. I'm not
expecting to be able to put all settings to max as I know that the newer
games require the multiple cores to process everything but compared to
what I am running games as now, it should be a huge improvement.

--

My blog - http://degeneratemoo.livejournal.com

'Moo' wrote:
| I will be putting off the upgrade for as long as possible as I simply
| couldn't afford to replace everything in the PC, for £140 I can get
| reasonable performance with the system I have for a while yet. I'm not
| expecting to be able to put all settings to max as I know that the newer
| games require the multiple cores to process everything but compared to
| what I am running games as now, it should be a huge improvement.
_____

I understand the need to hold on - my all-at-once upgrade was from a Pentium
2.4 GHz @ 3.2 GHz and RDRAM with FX5200 - which had to struggle with
Civilization IV B^)

And which was an upgrade from a Pentium III 1 GHz @ 1.3 GHz with a GeForce
TI 4000

And which was an upgrade from a Celeron 633 MHz @ 950 MHz with a GeForce 2

And which was an upgrade from a Celeron 300a @ 450 MHz with an Fire 1000GL

And which was an upgrade from a Pentium Classic 90 MHz @ 120 MHz with an ATI
Rage??

Only the cases have remained the same since the Celeron 300a system.

With the all-at-once upgrades I usually end up having two systems running;
and an alternative when tweaks go wrong B^(

Phil Weldon

"Moo" wrote in message
...
| Phil Weldon wrote:
| 'Moo' wrote:
| | Erm, what?
| |
| | I don't have an SLI board, just a standard AGP hence I will need to
| | upgrade everything.
| |
| | I know that the 7950 doesn't do DX10, if I wanted DX10 I would upgrade
| | everything and get the appropriate card for that, as it stands I'm
just
| | trying to get more time out of my current PC for as little as
possible.
| _____
|
| Since you will eventually be updating the entire system, it is even more
| important to get the best value DX9 AGP board that will meet your gaming
| requirements. If you'd share your time frame for a more complete upgrade
I'm
| sure you'll get a lot more suggestions. (A basic SLI nVidia
motherboard, a
| Core 2 Duo CPU, and 2 GBytes of DDR2 memory costs less than a 8800 GT
PCI-E
| video card - I've got an Intel 4300 Core 2 Duo @ 2.7 GHz, two GBytes
| DDR2-1066 memory, an EVGA 680i SLI motherboard, and an EVGA 8800 GTS 320
| MByte PCI-E card that gets about 10,000 on 3DMark06 with overclocking.
A
| reasonable upgrade path could include a faster Core 2 Duo [or Core 2 Duo
| Quad] and/or a second 8800 GTS or two 8800 GT cards.
|
| Raw frame rates don't exactly equal increased gaming capability across
the
| board of requirements. For the same game used in a FPS benchmark, but
with
| increased on screen and off screen complexity, extra FPS benchmark score
on
| the original scripted game benchmark with less complex will not
translate
| directly into increased game performance. A better basis to choose is a
| description of, or experience with a particular display adapter and a
| particular game.
|
| Phil Weldon
|
|
| "Moo" wrote in message
| ...
| | Phil Weldon wrote:
| | 'Moo' wrote:
| | | Nice, thanks for that... might be investing in one of those!
| | _____
| |
| | Well, the FPS rating is not exactly linear as a comparison
benchmark.
| After
| | all, who can SEE 1000 frames per second (or 100). At most, an FPS
| rating
| | may indicate the EXCESS CAPACITY of a CPU/Memory System/Graphics
System)
| for
| | displaying a scripted section of a game. An EXCESS CAPACITY that
may or
| may
| | not ever be called on. [Not to mention that 160% does not even seem
to
| be a
| | number that can be extracted from the cited page!]
| |
| | Benchmarks are of limited value, ESPECIALLY when improperly
interpreted.
| | And as long as "when I do decide to upgrade the PC completely I'm
gonna
| have
| | to upgrade every component has technology has moved on drastically"
then
| | surely the non DX10 capable 7950 GT would not be in the cards. What
SLI
| | gives you is a choice - install one card, and, if game play is not
| | satisfying to you, then adding a second, identical card is an easy
| upgrade.
| |
| | Unless you are going for bragging rights evaluation of game play is
much
| | more useful than comparing benchmarks.
| |
| | Phil Weldon
| |
| |
|
| I will be putting off the upgrade for as long as possible as I simply
| couldn't afford to replace everything in the PC, for £140 I can get
| reasonable performance with the system I have for a while yet. I'm not
| expecting to be able to put all settings to max as I know that the newer
| games require the multiple cores to process everything but compared to
| what I am running games as now, it should be a huge improvement.
|
| --
|
| My blog - http://degeneratemoo.livejournal.com