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.
It's not the problem. Something is up with that config and the TNT2 loves it or the Max hates it.
Test System 1: G400 Max, PIII 600, 384 Mb RAM, Asus P2B-F, Aureal Vortex 2
Test System 2: Guillemot TNT2 Ultra, PIII 500, 384 Mb RAM, Asus P2B-D, Turtle Beach Vortex 2
Normally, using realitively easy settings, the G400/PIII 600 based system thrashes the TNT2U/PIII 500 system in Quake III. Here are some marks using Normal settings, 800x600, 16-bit/16-bit, everything else at default:
G400: 62.8 fps
TNT2U: 49.7 fps
Using the test config posted here, the situation reverses dramatically:
G400: 40.9 fps
TNT2U: 57.2 fps
I did no tweaking, nor did I attempt to investigate further.
I used the drivers associated with PowerDesk 5.41 and the new TurboGL with the G400 Max. With the Xentor32, I used Guillemot's 2.40 drivers. Both systems had fresh installs of Quake3, and relatively fresh installs of Windows 98se.
Both systems have 3Comm NIC's. I'm using the drivers that installed with Windows. No goofy 3Comm utilities were installed.
Right we are both running Win98SE in 32bit colour desktop.
Hex res :1024*768
My res :1152*864
here are Hex's IRQ's
Setting Hardware using the setting
======= ==========================
00 System Timer
01 PC/AT Enhanced Keyboard (101/102-Keys)
02 Programmable interupt controller
04 Communications Port (Com1) (Note : Used by External Modem)
05 Creative SB16 Emulation
06 Standard Floppy Disk Controller
07 ECP Printer Port (LPT1)
08 System CMOS/real time clock
09 SCI IRQ used by ACPI bus
10 VIA Tech PCI to USB Universal Host Controller
10 IRQ Holder for PCI Steering
10 VIA Tech PCI to USB Universal Host Controller
11 ACPI IRQ Holder for PCI IRQ Steering
11 IRQ Holder for PCI Steering
11 Matrox Millenium G400 DualHead Max - English
12 ACPI IRQ Holder for PCI IRQ Steering
12 IRQ Holder for PCI Steering
12 Creative SB Live!
14 Via Bus Master PCI Controller (Ultra DMA)
14 Primary Bus Master IDE Controller (dual fifo)
15 Via Bus Master PCI Controller (Ultra DMA)
15 Secondary Bus Master IDE Controller (dual fifo)
hmm just read that post concerning how my cfg seems tnt2u biased. Thats quite wird because its basically the same as i used for my g200 only in a higher resolution, any chance somebody like jwilliams who is getting good framerates with high texture on and Athlon/Max system could post their cfg.
If you do post your cfg REMEMBER TO OPEN IT UP AND DELETE THE CD-KEY IN IT FIRST. I can't stress this enough because if somebody does get your Q3 CD-Key you are going to have real problems playing online. Cfg's without a CD-Key in them have the CD-Key as 123456789 in them.
I didn't mean for that to sound like a value judgement. But, given that my findings coincide with your findings, I can think of three scenarios (because that's all I can think of at the moment):
Something about these settings is making the G400 choke.
Something about these settings has enhanced the TNT2's performance.
You and I have made a similar error or our systems have similar conflicts resulting in the almost identical results.
Given the third scenario is *not* true, we have most likely eliminated the Athlon as a variable.
Paul
paulcs@flashcom.net
[This message has been edited by paulcs (edited 09 January 2000).]
OK. Both Fyrespray and I lowered out texture sliders by one. Heres what we got.
Fyrespray : 58.7
Hex : 51.3
I noticed just over a 10 fps gain, while Fyre just got a 5fps increase.
Also to give a description of what actually gameplay is like, when the fps drop, it isnt a smooth transition, its normally quite sudden, either reducing to a slight pause or stuttering, also ive noticed random HD accesses mid game (audible HD accessing sometimes, others rapid led flashing), doesnt apear to be loading textures or sounds, they just happen, they cause the game to pause for around half a second each time, maybe this is a symptom of a problem? I have a 9 Gig IBM 7200Rpm HD Drive with DMA switched on.
---Hex
[This message has been edited by Hex (edited 09 January 2000).]
"There's no such thing as a stupid question, just stupid people"
That's weird, the only time I have ever experienced that is when I had a bad stick of RAM that would occassionally flake-out, and not be recognized by windows, therefore making me run on 64MB. Hex, hit your windows key and break key at the same time, and see if your memory amount is listed properly under general. If it is okay, hit ctrl/alt/del shutdown PDESK.EXE (you may have to c/a/d and shut it down several times before it totally shuts off), then run your benchmark again. Also, if this helps, you may want to switch to PD 5.30 with TGL.
Ok.
Windows shows my full amount of ram ok.
128.00 Mb. Closing pdesk.exe (powerdesk) gave me no improvement. In fact quake 3 was unable to show all textures with it closed. Ive tried the 5.30 drivers before, same experiences with those.
---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"
No we dont mind at all. Your posts are keeping me sane and stopping me from ripping out all the components in this sytem and smashing them into tiny pieces :P
All I hope is that this can be sorted cause its getting a bit frustrating now.
----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"
That is very strange, I get no slowdowns like you describe at all, and I have a P3. I also wonder about which drivers the TNT2U has, and what are the settings for them at? I know that there are tweaks that allow for 32bit textures to be dithered down, and reduced textures, I am curious if he doesn't have those tweaks enabled, thus giving a false comparison.
And you know what would be a better comparison is to take and swap out vide cards with each other to see which is faster in each particular system, that would give a good indicator of which is which.
Also, another thing I noticed is that trilinear filtering is enabled in this configuration, the TNT2U doesn't properly support trilinear filtering, so you will get false results there, too. Disable trilinear filtering for both to do a proper compare.
I'm using 3.56 detonator driver with the card i havn't tweaked it just put it in loaded up the drivers and let it get on with it, the puzzleing thing here is that in an earlier post JWilliams said that he was benchmarking at 77 FPS odd with an Athlon 750, now Athlon 750 isn't that much faster than a 700, one of the reasons being they slowed the cache down slighty on them. There must be some reason Hex's system is running so slow compared to his. TBH its quite possible that we havn't got the chip set up properly or forgot o change a setting in the bios or somthing else stupid that has been overlooked, any suggestions from Athlon owners as to bios settings or anything we might have missed on install would be welcome.
Comment