Welcome, Unregistered.

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.

Closed Thread
Old 28-Sep-2004, 19:49   #1
trinibwoy
Meh
 
Join Date: Mar 2004
Location: New York
Posts: 9,809
Default 3DMark05 scores courtesy the Inq

Inquirer

More driver magic goodness

Quote:
Originally Posted by The Inquirer
With current drivers, Nvidia will score 5000 points with the 6800 Ultra while the X800XT PR will score only 4500 with 4.9 drivers. With the new ATI driver codenamed 8.07, ATI score will jump to an incredible 5800. All these scores are on a 3.4GHz Pentium 4 machine. The ATI driver will be introduced as the Catalyst 4.10 and we learned it should get its WHQL very soon.

The Geforce FX 5950 Ultra will score a modest 1270 while the Radeon 9600XT will provide 1570 marks. The mainstream card kicks the former high end in its performance nuts.

The Geforce 6800 GT will score 4500.

You will be able to see many more scores on the 29th of this month at 3PM CET when Futuremark lifts the non disclosure agreement (NDA) veil.
__________________
What the deuce!?
trinibwoy is offline  
Old 28-Sep-2004, 19:57   #2
PowerK
Member
 
Join Date: Jun 2004
Location: Seoul, Korea
Posts: 312
Default

I wonder if Inq meant XT as in plain non-PE XT, or XT PE. And 1300 points improvement going from Catalyst 4.9 to Catalyst 4.10 ? Hmm.................. :?
PowerK is offline  
Old 28-Sep-2004, 20:07   #3
Tim Murray
chaos dunk
 
Join Date: May 2003
Location: Mountain View, CA
Posts: 3,274
Default

... why did it jump 1300 points? that seems, oh, I don't know, INSANE.

filtering tricks? am I allowed to say the naughty "shader replacement" word?
Tim Murray is offline  
Old 28-Sep-2004, 20:20   #4
trinibwoy
Meh
 
Join Date: Mar 2004
Location: New York
Posts: 9,809
Default

Quote:
Originally Posted by PowerK
I wonder if Inq meant XT as in plain non-PE XT, or XT PE. And 1300 points improvement going from Catalyst 4.9 to Catalyst 4.10 ? Hmm.................. :?
I think the 'PR' was supposed to be 'PE'. And not only is it 1300 points but it's a 29% increase
__________________
What the deuce!?
trinibwoy is offline  
Old 28-Sep-2004, 20:21   #5
Bouncing Zabaglione Bros.
Regular
 
Join Date: Jun 2003
Posts: 6,170
Default

Quote:
Originally Posted by PowerK
I wonder if Inq meant XT as in plain non-PE XT, or XT PE. And 1300 points improvement going from Catalyst 4.9 to Catalyst 4.10 ? Hmm.................. :?
Looks like it's a typo - "XT PR" should be "XT PE", and it makes sense if you are comparing it to the 6800U.
Bouncing Zabaglione Bros. is online now  
Old 28-Sep-2004, 20:26   #6
Miksu
Member
 
Join Date: Mar 2003
Location: Finland
Posts: 943
Default

8.07 is the hotfix-driver which was released little earlier. It will be interesting to see which sites will be using these drivers in their tests and which will use older ones..
__________________
Mikael Koskinen blog: .NET Programming, Windows Phone Development, Software Architecture
Miksu is offline  
Old 28-Sep-2004, 20:43   #7
Mendel
Mr. Upgrade
 
Join Date: Nov 2003
Location: Finland
Posts: 1,335
Default

http://www2.ati.com/drivers/hotfix/w...07-018380E.exe 8)
Mendel is offline  
Old 28-Sep-2004, 20:44   #8
trinibwoy
Meh
 
Join Date: Mar 2004
Location: New York
Posts: 9,809
Default

Does this hotfix magic work on R3xx also?
__________________
What the deuce!?
trinibwoy is offline  
Old 28-Sep-2004, 21:03   #9
IbaneZ
Member
 
Join Date: Apr 2003
Posts: 743
Default

Hotfix indeed.
IbaneZ is offline  
Old 28-Sep-2004, 21:35   #10
Pete
Moderate Nuisance
 
Join Date: Feb 2002
Posts: 4,653
Default

1. If it was Fudo who penned this, he probably meant to say XT PR, for obvious (availability) reasons.

2. Baron, "shader replacement" sounds so uncivilized. ATi much prefers the term "A.I." :P

3. 30%! I guess that makes nV's purported '05 boost with Rev's FM-approved NDA drivers look a little weak, by comparison. Does 3DM05 have a more comprehensive dev mode, to perhaps catch cheating--er, unusually intelligent drivers?
Pete is offline  
Old 28-Sep-2004, 21:41   #11
Dave Baumann
Gamerscore Wh...
 
Join Date: Jan 2002
Posts: 12,956
Default

Hint: the performance only affects 256MB boards and only AGP (i.e. PCIe were there or thereabouts already)
__________________
Expand. Accelerate. Dominate.
Tweet Tweet!
Dave Baumann is offline  
Old 28-Sep-2004, 21:47   #12
jvd
Naughty Boy!
 
Join Date: Feb 2002
Location: new jersey
Posts: 12,731
Send a message via AIM to jvd
Default

Quote:
Originally Posted by DaveBaumann
Hint: the performance only affects 256MB boards and only AGP (i.e. PCIe were there or thereabouts already)
tweaking the memory controller ? I wonder if it is the memory controller if we will see bigger performance gains in other games .
__________________
Freexbox 360 !!!
Free Psp!
jvd is offline  
Old 28-Sep-2004, 21:49   #13
pino
Member
 
Join Date: Dec 2003
Location: Naples, Italy
Posts: 109
Default

Quote:
Originally Posted by DaveBaumann
Hint: the performance only affects 256MB boards and only AGP (i.e. PCIe were there or thereabouts already)
so 9800XT too ?
pino is offline  
Old 28-Sep-2004, 21:52   #14
cloudscapes
Member
 
Join Date: Aug 2004
Location: Montreal, Canada
Posts: 451
Default Re: 3DMark05 scores courtesy the Inq

Quote:
The ATI driver will be introduced as the Catalyst 4.10 and we learned it should get its WHQL very soon.
*confused* 4.9 < 4.10? If so, it should really be 4.09 to not confuse ATI noobs like myself.
cloudscapes is offline  
Old 28-Sep-2004, 21:54   #15
Dave Baumann
Gamerscore Wh...
 
Join Date: Jan 2002
Posts: 12,956
Default

4.10: 4 = year (2004), 10 = month (October)
Thats the catalyst release number. The drivers will have a driver number which is normally sequential (i.e. these drivers are the 8.07's).
__________________
Expand. Accelerate. Dominate.
Tweet Tweet!
Dave Baumann is offline  
Old 28-Sep-2004, 21:55   #16
jvd
Naughty Boy!
 
Join Date: Feb 2002
Location: new jersey
Posts: 12,731
Send a message via AIM to jvd
Default Re: 3DMark05 scores courtesy the Inq

Quote:
Originally Posted by cloudscapes
Quote:
The ATI driver will be introduced as the Catalyst 4.10 and we learned it should get its WHQL very soon.
*confused* 4.9 < 4.10? If so, it should really be 4.09 to not confuse ATI noobs like myself.
Yes , its the 10th release of the 4th year of catalyst . So its newer than the 9th release of this year
__________________
Freexbox 360 !!!
Free Psp!
jvd is offline  
Old 28-Sep-2004, 22:25   #17
mreman4k
Junior Member
 
Join Date: Jul 2003
Location: Raleigh, NC
Posts: 82
Send a message via AIM to mreman4k
Default

deja vu :?
__________________
EMunXT
ABIT AN8 Ultra (nForce4 Ultra) || AMD Athlon64 X2 3800+ || 2GB Patriot PC3200 DDR SDRAM || ATI All-in-Wonder Radeon X800 XL 256MB || Audigy 2 || 120GB & 80GB Seagate HDs || Lite-On DVD Drive & DVD RW || Antec P160WF Case || Antec TruePower 2.0 430W PSU || Samsung 204T (20.1") & Hitachi CML174SX (17") LCDs

I'm black, I'm still sexy, my computer is now sexier though.â„¢
mreman4k is offline  
Old 29-Sep-2004, 00:14   #18
Padman
Junior Member
 
Join Date: Jan 2003
Location: Netherlands
Posts: 30
Default

/rant on

VERSION NUMBERS ARE NOT FRACTIONS!!!!!!

/rant off
Padman is offline  
Old 29-Sep-2004, 00:53   #19
Reznor007
Member
 
Join Date: Feb 2002
Location: Norman, OK, USA
Posts: 535
Send a message via ICQ to Reznor007 Send a message via AIM to Reznor007 Send a message via MSN to Reznor007 Send a message via Yahoo to Reznor007
Default

Quote:
Originally Posted by jvd
Quote:
Originally Posted by DaveBaumann
Hint: the performance only affects 256MB boards and only AGP (i.e. PCIe were there or thereabouts already)
tweaking the memory controller ? I wonder if it is the memory controller if we will see bigger performance gains in other games .
The driver update mentions a memory allocation bug being fixed, so it probably has to do with that.
Reznor007 is offline  
Old 29-Sep-2004, 00:59   #20
Mintmaster
Senior Member
 
Join Date: Mar 2002
Posts: 3,779
Default

Quote:
Originally Posted by Pete
3. 30%! I guess that makes nV's purported '05 boost with Rev's FM-approved NDA drivers look a little weak, by comparison.
Well, remember this?
before
after

I wonder if ATI did some insane shader replacement or just found a little bug slowing them down like mad. R3xx and R4xx always had quite optimal shader execution, so it would surprise me if it was the latter.

EDIT: Reznor007, you may be right. The X700 benchmarks are full of results that suggest memory allocation problems.
Mintmaster is online now  
Old 29-Sep-2004, 02:38   #21
Smurfie
Member
 
Join Date: Apr 2004
Posts: 107
Default

Nah, I would think it's just putting things back into its status quo.

525/400 > 5800/5000. ATI still has some work to do.
Smurfie is offline  
Old 29-Sep-2004, 02:43   #22
FUDie
Member
 
Join Date: Sep 2002
Posts: 559
Default

Quote:
Originally Posted by Smurfie
Nah, I would think it's just putting things back into its status quo.

525/400 > 5800/5000. ATI still has some work to do.
What a silly statement. Do you also compare AMD and Intel chips by looking at clockspeeds? If ATI and NVIDIA were selling the same chips, you might have a valid point.

-FUDie
__________________
Ph.D. - Piled Higher and Deeper
FUDie is offline  
Old 29-Sep-2004, 02:51   #23
Smurfie
Member
 
Join Date: Apr 2004
Posts: 107
Default

Nope. Reason being, the 6800 and X800 pipelines are far more similar than the AMD and Intel CPU pipelines. The IPC count between 1 clock cycle with an AMD and Intel CPU is far more different than 1 clock cycle of a NV40 and R420 GPU.

Besides, fillrate still seems to be a major determinant, and those fillrates are directly tied to the clockspeeds. Until I see a many applications breaking that rule, it still stands.
Smurfie is offline  
Old 29-Sep-2004, 03:22   #24
FUDie
Member
 
Join Date: Sep 2002
Posts: 559
Default

Quote:
Originally Posted by Smurfie
Nope. Reason being, the 6800 and X800 pipelines are far more similar than the AMD and Intel CPU pipelines. The IPC count between 1 clock cycle with an AMD and Intel CPU is far more different than 1 clock cycle of a NV40 and R420 GPU.

Besides, fillrate still seems to be a major determinant, and those fillrates are directly tied to the clockspeeds. Until I see a many applications breaking that rule, it still stands.
So I guess you've forgotten about the free FP16 normalization in the NV40? Also, what if the benchmark is running in SM 3.0 for the NV40? Those are just some of the differences between the two architectures.

Your logic is flawed.

-FUDie
__________________
Ph.D. - Piled Higher and Deeper
FUDie is offline  
Old 29-Sep-2004, 03:43   #25
cloudscapes
Member
 
Join Date: Aug 2004
Location: Montreal, Canada
Posts: 451
Default

Quote:
Originally Posted by Padman
/rant on

VERSION NUMBERS ARE NOT FRACTIONS!!!!!!

/rant off
Okay okay sorry. My mistake. I've only been usign ATI drivers for like a month and a half. I had to find out eventually.
cloudscapes is offline  

Closed Thread

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

BB 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
3dmark05 & ATI - Sacrificing Image Quality for Speed? Ruined 3D Hardware, Software & Output Devices 94 07-Oct-2004 15:03
Higher 3DMark05 scores = Better Equipped for Future Games? Reverend 3D & Semiconductor Industry 57 07-Oct-2004 00:30
Reverend at The Pulpit #12 Reverend General Discussion 73 03-Oct-2004 10:13
3DMark05 and certain websites Reverend 3D & Semiconductor Industry 116 01-Oct-2004 10:46
Hype and review scores. Geeforcer Console Technology 67 05-Dec-2002 07:31


All times are GMT +1. The time now is 10:22.


Powered by vBulletin® Version 3.8.6
Copyright ©2000 - 2013, Jelsoft Enterprises Ltd.