A computer components & hardware forum. HardwareBanter

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

Go Back   Home » HardwareBanter forum » Video Cards » Nvidia Videocards
Site Map Home Register Authors List Search Today's Posts Mark Forums Read Web Partners

Squeezing every last point out of 3DMark2001SE



 
 
Thread Tools Display Modes
  #1  
Old February 13th 04, 07:54 AM
Kai Robinson
external usenet poster
 
Posts: n/a
Default Squeezing every last point out of 3DMark2001SE

I write this at 7:42am - and i still havent gone to sleep yet. I've been
spending hours tweaking this system in hopes of getting just that little bit
more out of my system. After changing my Ram timings to 2-2-5-2, enabling
4-way bank interleaving and upping the pulse width to 8. Clocking the card
to 240/490 (and bearing in mind this is 5ns stuff - thats 400Mhz top speed)
and running the processor at 1604Mhz (153 x 10.5 Multiplier) - i finally got
the result i was after. Now it may not be much - but to me this is a huge
amount - especially for a 'budget' GeForce 3 Ti200.

Previously - the best score i got at the same settings (except the Ram
timings), was 8158. With the ram timings enabled, i'm getting 8767 - with
most of the difference being made in the Game benches, Single Texturing Fill
rate (766.5 vs 746.4 MTexels/sec) and 1 light High Polygon count (29.0 vs
27.9 MTriangles/sec)

Every other score was virtually identical.

Thing is though - can i do better?

Anyone else managed to get a score this high with one of these cards?

Kai



  #2  
Old February 13th 04, 09:28 AM
Clock´n Roll
external usenet poster
 
Posts: n/a
Default


"Kai Robinson" skrev i en meddelelse
...
I write this at 7:42am - and i still havent gone to sleep yet. I've been
spending hours tweaking this system in hopes of getting just that little

bit
more out of my system. After changing my Ram timings to 2-2-5-2, enabling
4-way bank interleaving and upping the pulse width to 8. Clocking the card
to 240/490 (and bearing in mind this is 5ns stuff - thats 400Mhz top

speed)
and running the processor at 1604Mhz (153 x 10.5 Multiplier) - i finally

got
the result i was after. Now it may not be much - but to me this is a huge
amount - especially for a 'budget' GeForce 3 Ti200.

Previously - the best score i got at the same settings (except the Ram
timings), was 8158. With the ram timings enabled, i'm getting 8767 - with
most of the difference being made in the Game benches, Single Texturing

Fill
rate (766.5 vs 746.4 MTexels/sec) and 1 light High Polygon count (29.0 vs
27.9 MTriangles/sec)

Every other score was virtually identical.

Thing is though - can i do better?

Anyone else managed to get a score this high with one of these cards?


You should ask:"Anyone else managed to get a score this high with a 1.6Ghz
CPU?"
3Dmark2001 in my oppinion is more a CPU test than a GPU test.



  #3  
Old February 13th 04, 12:57 PM
~misfit~
external usenet poster
 
Posts: n/a
Default

Clock´n Roll wrote:
"Kai Robinson" skrev i en meddelelse
...
I write this at 7:42am - and i still havent gone to sleep yet. I've
been spending hours tweaking this system in hopes of getting just
that little bit more out of my system. After changing my Ram timings
to 2-2-5-2, enabling 4-way bank interleaving and upping the pulse
width to 8. Clocking the card to 240/490 (and bearing in mind this
is 5ns stuff - thats 400Mhz top speed) and running the processor at
1604Mhz (153 x 10.5 Multiplier) - i finally got the result i was
after. Now it may not be much - but to me this is a huge amount -
especially for a 'budget' GeForce 3 Ti200.

Previously - the best score i got at the same settings (except the
Ram timings), was 8158. With the ram timings enabled, i'm getting
8767 - with most of the difference being made in the Game benches,
Single Texturing Fill rate (766.5 vs 746.4 MTexels/sec) and 1 light
High Polygon count (29.0 vs
27.9 MTriangles/sec)

Every other score was virtually identical.

Thing is though - can i do better?

Anyone else managed to get a score this high with one of these
cards?


You should ask:"Anyone else managed to get a score this high with a
1.6Ghz CPU?"
3Dmark2001 in my oppinion is more a CPU test than a GPU test.


My GF4 ti4200 with a CPU running at 2.1 GHz gets 12,000.

My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets 2,500.

While the CPU does have an influence on the 3DMark score it isn't anywhere
near as significant as the GPU.
--
~misfit~


  #4  
Old February 13th 04, 08:05 PM
cowboyz
external usenet poster
 
Posts: n/a
Default


"~misfit~" wrote in message
...


My GF4 ti4200 with a CPU running at 2.1 GHz gets 12,000.

My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets 2,500.

While the CPU does have an influence on the 3DMark score it isn't anywhere
near as significant as the GPU.
--
~misfit~



surely these figures are a typo. 2500? or is that 12500?


  #5  
Old February 13th 04, 09:53 PM
Clock´n Roll
external usenet poster
 
Posts: n/a
Default


"cowboyz" skrev i en meddelelse
...

"~misfit~" wrote in message
...
My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets

2,500.

Right.... :-D


  #6  
Old February 14th 04, 12:36 AM
~misfit~
external usenet poster
 
Posts: n/a
Default

Clock´n Roll wrote:
"cowboyz" skrev i en meddelelse
...

"~misfit~" wrote in message
...
My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets
2,500.


Right.... :-D


Sorry, made a mistake on that figure, see my reply to cowboyz.
--
~misfit~


  #7  
Old February 14th 04, 12:35 AM
~misfit~
external usenet poster
 
Posts: n/a
Default

cowboyz wrote:
"~misfit~" wrote in message
...


My GF4 ti4200 with a CPU running at 2.1 GHz gets 12,000.

My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets
2,500.

While the CPU does have an influence on the 3DMark score it isn't
anywhere near as significant as the GPU.
--
~misfit~



surely these figures are a typo. 2500? or is that 12500?


Sorry, it was a mistake. The 2,500 (2,433 actually) was what she got with
her old GF2MX400/64. With the FX5200/128 she still only gets 5,585 though. I
must read my records more carefully. :-). That's on a 2500+ Barton clocked
to 3200+ speed (200MHz FSB) with 512MB RAM.

(I'm a bit 'anal' I guess the Americans would say, I rigorously benchmark
and record every build or change I make for future reference).

The FX5200/128 (non-ultra, Gigabyte) sucks quite a bit but it was only
around $NZ120. Anything half-way decent was around $NZ300 (I bought it for
her christmas present and didn't/don't have that sort of money). Although it
benchmarks quite badly it seems to run better than the benchmark indicates.
All our games are playable on it. It even plays the Far Cry demo with only
slight hesitation (Thanks BTW) *Far* better than the GF2 that is now in our
mp3 machine (and gets 1,937 marks on a celly 600 at 900MHz, 256MB SD-RAM).

All using XP Pro.
--
~misfit~


  #8  
Old February 14th 04, 05:52 AM
Darthy
external usenet poster
 
Posts: n/a
Default

On Sat, 14 Feb 2004 12:35:19 +1300, "~misfit~"
wrote:

cowboyz wrote:
"~misfit~" wrote in message
...


My GF4 ti4200 with a CPU running at 2.1 GHz gets 12,000.

My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets
2,500.

While the CPU does have an influence on the 3DMark score it isn't
anywhere near as significant as the GPU.
--
~misfit~



surely these figures are a typo. 2500? or is that 12500?


Sorry, it was a mistake. The 2,500 (2,433 actually) was what she got with
her old GF2MX400/64. With the FX5200/128 she still only gets 5,585 though. I
must read my records more carefully. :-). That's on a 2500+ Barton clocked
to 3200+ speed (200MHz FSB) with 512MB RAM.


Thats about right...

GF2s are in the 2500 range... GF3s are in the 5000~7000s...

(I'm a bit 'anal' I guess the Americans would say, I rigorously benchmark
and record every build or change I make for future reference).


Many of us do and should... so we know what were getting.


--
Remember when real men used Real computers!?
When 512K of video RAM was a lot!

Death to Palladium & WPA!!
  #9  
Old February 14th 04, 02:45 AM
Darthy
external usenet poster
 
Posts: n/a
Default

On Sat, 14 Feb 2004 00:57:15 +1300, "~misfit~"
wrote:

You should ask:"Anyone else managed to get a score this high with a
1.6Ghz CPU?"
3Dmark2001 in my oppinion is more a CPU test than a GPU test.


My GF4 ti4200 with a CPU running at 2.1 GHz gets 12,000.

My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets 2,500.

While the CPU does have an influence on the 3DMark score it isn't anywhere
near as significant as the GPU.


That is WAY horrible for a 5200... there should be MORE issues than
just the crappy 5200... thats SLOWER than the GF3... about the same as
a GF2mx400 ... which on paper is what a 64bit 5200 can perform at...
roughly.

A $75 Ti4200 would get her up to at least 9000.


--
Remember when real men used Real computers!?
When 512K of video RAM was a lot!

Death to Palladium & WPA!!
  #10  
Old February 14th 04, 10:00 AM
~misfit~
external usenet poster
 
Posts: n/a
Default

Darthy wrote:
On Sat, 14 Feb 2004 00:57:15 +1300, "~misfit~"
wrote:

You should ask:"Anyone else managed to get a score this high with a
1.6Ghz CPU?"
3Dmark2001 in my oppinion is more a CPU test than a GPU test.


My GF4 ti4200 with a CPU running at 2.1 GHz gets 12,000.

My g/f's FX5200 with a CPU at 2.2 GHz and twice the L2 of mine gets
2,500.

While the CPU does have an influence on the 3DMark score it isn't
anywhere near as significant as the GPU.


That is WAY horrible for a 5200... there should be MORE issues than
just the crappy 5200... thats SLOWER than the GF3... about the same as
a GF2mx400 ... which on paper is what a 64bit 5200 can perform at...
roughly.

A $75 Ti4200 would get her up to at least 9000.


Except I can't buy one new here in NZ. Tried for a week.
--
~misfit~


 




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Balance Point, AGP Overclocking David B. Overclocking 6 April 19th 05 01:42 PM
Passmark Performance Test, Division, Floating Point Division, 2DShapes @(none) General 0 August 19th 04 11:57 PM
Wireless LAN - access point required? Erik Hegeman General 3 June 2nd 04 10:10 AM
Any point in using PC4000 memory? Barend Overclocking 9 January 21st 04 12:01 AM
Have A7A266; any point installing Zalman north bridge heat sink rstlne Asus Motherboards 0 July 28th 03 02:56 PM


All times are GMT +1. The time now is 02:53 PM.


Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 HardwareBanter.
The comments are property of their posters.