Why don't you try the Ziff-Davis benchmarks (ZDBOP)? They have suites that will test each of the subsystems as well as the system as a whole.
Announcement
Collapse
No announcement yet.
G400MAX beatan by a TNT2U in Q3??????????
Collapse
X
-
Sandra doesn't fully support Athlon processors. This is from SiSoft's site:
"Next Release Timeframe: The next version of Sandra will be out shortly. The release was delayed in order to fully support (as well as test on) AMD Athlon and Intel PIII Coppermine processors as well as other new technologies such as DirectX7.
"The benchmarks have been changed to support 3DNow! Enhanced instruction set, AMD Enhanced MMX instruction set, 7th generation processors, RDRAM & PC133 SDRAM in order to return more meaningful indexes. However, being synthetic benchmarks any indexes may or may not tally with real-life performance."
Have you guys tried the Quake2 Crusher demo. There are timedemos available for Half Life and UT as well. You might want to figure out if this is a Quake3 related issue, or if it transmits to other real world tests as well. 3DMark 99 Max (*not* 2000!) might be a good test as well.
The G400 should clean the TNT2's clock in 3DMark 99 Max. If it doesn't, we know something is really wrong. Also, there are scores all over the forums to compare to.
The Crusher timedemo, I think, is still a good OpenGL indicator.
Paul
paulcs@flashcom.net
Comment
-
Paulcs, Yes that's right about Sandra (I have reg Pro ver), but it's all relative. I can set mine to what ever they may have... if he's performing subpar on it, then that would be a good reference point to work from."Be who you are and say what you feel, because those who mind don't matter, and those who matter don't mind." -- Dr. Seuss
"Always do good. It will gratify some and astonish the rest." ~Mark Twain
Comment
-
I agree, there is a lot of memory latency on the current Athlon MB's. However, there is Super Bypass, if the MB has one of the newer AMD chipsets on board. It's supposed to reduce the latency by 15-25%, and give a 5-15% performance increase in real-life. Any Athlon owners out there with a Rev. C, Stepping 5 chip on their board?
Gigabyte P35-DS3L with a Q6600, 2GB Kingston HyperX (after *3* bad pairs of Crucial Ballistix 1066), Galaxy 8800GT 512MB, SB X-Fi, some drives, and a Dell 2005fpw. Running WinXP.
Comment
-
Most of the SlotA MB out right now don't! Mine's a stepping 4(
"Be who you are and say what you feel, because those who mind don't matter, and those who matter don't mind." -- Dr. Seuss
"Always do good. It will gratify some and astonish the rest." ~Mark Twain
Comment
-
The only 5's I've heard about so far are the latest K7M's, and some of the scary, no-name companies. I think I read about a BCM board with 5's.
Gigabyte P35-DS3L with a Q6600, 2GB Kingston HyperX (after *3* bad pairs of Crucial Ballistix 1066), Galaxy 8800GT 512MB, SB X-Fi, some drives, and a Dell 2005fpw. Running WinXP.
Comment
-
Hi
I have a system similar to Firespray's
Cel400@500, 128 meg PC100, Sblive but I have a G400 MAX.
Here are some score and tought:
using Q3DEMO and tour config files
PD5.30 installed (previous benchmark show no difference between PD5.41 and PD5.30 on my system).
So
Tri:40,3
Bi:51 and with text down one notch :54
So as you can see:
1) I get the same score than Hex, which is highly anormal since he's on a 700 Mhz Athlon and I'm on a Celeron@500, knowing that the G400 scale with computer power, there's something wrong here.
2) The TNT2Ultra is better than the g400 on an around 500 Mhz Celeron system, unless Firespray did o/c his TNT2U, because by o/c my MAX I could get around his score. Unless the 50 Mhz diff between our system do make all the difference, but I don't think so.
3)I did some other bench (too lazy to put them here) showing in Q3 that full texture and trilinear brings a drop in performance around 18% as opposed to bilinear.(Nvidia surely cheat here, doesn't surprise me...)
If you drop the texture one notch, the drop is around 4-5%, that support what has been said before that max texture eats a lot of mem and trilinear choke the card in this situation.
But to conclude (I know it's long...).
Hex have definitly a problem with his setting.
[This message has been edited by jackzod (edited 10 January 2000).]Athlon64 4800+
Asus A8N deluxe
2 gig munchkin ddr 500
eVGA 7800 gtx 512 in SLI
X-Fi Fatality
HP w2207
Comment
-
So has anyone out there got anything i can try? PLEASE! Im going frickin' nuts here.
HEEEEEEELP!
----Hex
------------------
"There's no such thing as a stupid question, just stupid people"
"There's no such thing as a stupid question, just stupid people"
Comment
-
Hex & Fry:
Aren't you listening? nVidia DOESN'T DO TRILINEAR FILTERING. At all. Period. End of story. The card doesn't support it and neither do any current drivers.
The G400 _DOES_ DO TRILINEAR FILTERING.
Therefore, with trilinear turned ON, the TNT will ALWAYS outperform the G400, because the G400 is taking a 30+% performance hit due to trilinear filtering.
Turn off Trilinear and you will discover the G400 walks all over the TNT2.
End of story.
- Gurm
------------------
Listen up, you primitive screwheads! See this? This is my BOOMSTICK! Etc. etc.The Internet - where men are men, women are men, and teenage girls are FBI agents!
I'm the least you could do
If only life were as easy as you
I'm the least you could do, oh yeah
If only life were as easy as you
I would still get screwed
Comment
-
Ok. Forget the TNT2u VS G400Max for a moment, im more concerned now at why my system apears to be so slow. Im roughly on average 7 fps slower than a Celery 550. Im using a 700 athlon and am experiencing lower fps (even with Triliniar OFF i still am being out performed by the TNT2u + Cellery system). Please can anyone provide me with possible problems and maybe ways of fixing them. Cheers
----Hex
------------------
"There's no such thing as a stupid question, just stupid people"
"There's no such thing as a stupid question, just stupid people"
Comment
-
Check your Mail!!
I have a similar Athlon system, I want to get my sytem faster too... Please
Here is my system:System Information:
Processor Type: AMD-K7(tm) Processor
Processor Speed: 500MHz
Processor Caps: Enhanced 3D Now!/MMX
Total System RAM: 128Mb Dimm PC100
CPU/Memory Bandwidth: 184 MB/s (VERY SHITE)
FPU/Memory Bandwidth: 152 MB/s
Motherboard: MS-6167 (AMD751)
Sound Card: Creative Labs Sound Blaster PCI128
Graphics Card: Maxtrox G400 MAX
RESULTS TEST 2
3DMark Result: 2074 2538
Game 1 - Helicopter
Low: 43.0 51.0
Medium: 25.8 34.2
High: 8.1 14.5
Game 2 - Adventure
Low: 47.8 58.0
Medium: 30.3 34.0
High: 17.9 19.8
Fill Rate (Single-Texturing): 226.3 17871.9
Fill Rate (Multi-Texturing): 236.9 268.7
High Polygon Count (1 Light): 1587 2248
High Polygon Count (4 Light): 1417 1910
High Polygon Count (8 Light): 1225 1572
8MB Texture Rendering Speed: 154.1 224.8
16MB Texture Rendering Speed: 146.1 217.5
32MB Texture Rendering Speed: 115.8 156.9
64MB Texture Rendering Speed: 103.1 68.1
Bump Mapping (Emboss, 3-pass): 61.3 98.2
Bump Mapping (Emboss, 2-pass): 84.1 126.8
Bump Mapping (Emboss, 1-pass): 156.8 226.5
Bump Mapping (Enviroment): 73.7 89.8
Internal 3D CARD
Name: Matrox Millennium G400 DualHead Max - English
Driver Name: G400DD32.DLL
Driver Version: 4.11.01.1410 (Not Certified)
Total Video Memory: 32697KB
Total Texture Memory: 28809KB
Bus: AGP
IRQ line: 9
Comment
-
Out of Site, Out of Mind.
=========================
Just installed Win2k, now i cant even run Quake 3 :-) So i can stop griping about my slow quake 3 fps for a few months until new drivers come out =P. Win2k the way to go.
----Hex
------------------
"There's no such thing as a stupid question, just stupid people"
"There's no such thing as a stupid question, just stupid people"
Comment
Comment