hey guys, first off, buckeye, make sure your v3 doesn't get too hot with that heatsink on the back- overclock9 knows all too much about frying cards that way. also, getting more ram will smooth out nfs3- i've run it on two computers with 128 mb, and they run super smooth. theman- i think rwc is right, i think you need the full version of quake2, and if you get it, also get the 3.20 patch for it. kinda sucks, but i think that's what you need.
now, for the benchs. system info is 300a @ 472 on an abit zm6 mobo, western digital caviar 13 gb ultra ata/66 hd, 128 mb of pc-100 ram, 40x cdrom, 286 1.44 mb floppy drive (it still works like everyone else's!), yahmaha wavetable cheapo 64 bit soundcard (pci), TNT O/C to 114/122, a us robotics 56k isa modem, and a 10/100 baset nic (pci). i was also running a 20" box fan off of the side of my case to keep everything at an agreeable temp (TNT was barely warm, and my cpu temp was 84 and case 77 after i was done!) i only had one solid lock running this o/c'ed, and never had another glitch. anyhow, for the stuff... quake2 running the 3.20 patch, low sound quality (which i play with), on fov 110, using the launcher.cfg file, motherboard monitor lite 2.2 and powerstrip running the whole time.
quake2 16 bit timedemo massive crusher
640x480 85.7 66.4 42.5
800*600 75.1 61.3 41.5
960*720 56.6 49.8 38.2
1024*768 50.4 45.2 36.2
1152*864 40.6 37.2 31.2
quake2 32 bit
640*480 75.9 61.9 41.4
(you get 52.4 47.0 37.3
the idea) 36.5 33.5 29.1
31.6 29.4 25.9
24.7 23.0 20.2
3833 3dmarks, 4551 cpu marks
3924 3dmarks, 4567 cpu marks
viper racing, with everything on and normal sound
512*384- 55.2, 640*480- 52.6, 800*600- 51.4, 1024*768- 54.4
viper racing was really weird, seeing as that my second highest score was in the highest res, and it ran better too! it was smooth, and looked a lot better too, so i know what res i'm using when i play. anything lower than 800*600 looked like crap- don't bother playing it. if you want to know about the q2 tweaks, talk to overclock9- he put them on, i don't know what they were besides that they help. my biggest differences were from post-800*600 and after 1024*768. all in all, i thought it ran great while benching, and the box fan really helped. as i get more benchs (couldn't get the one for sin to work, and i have to get shogo from vidfreek, nfs3?), i will post them, and i hope you enjoyed these ones. another thing is that when my friend isetnofret was doing his, he did them @464, and when he jumped to 472, his fps actually dropped! don't have any idea on that one, but i thought you guys might find that interesting. anyhow, good luck. later
why can i never think of anything witty to put in these?
Hey, Buckeye yeah Lavajumper is right I fried my old TNT by putting a heatsink on the back of it. You may want to monitor that thing closely, but as long as its not touching any metal points in the back you are safe unlike mine. I just have a 486 fan on my V3's stock heatsink. I'm a little weary about putting it past 190mhz which is what I have it at now. Don't want to have to get another video card like I had to after I fried my TNT. Oh and just to clear something up, some people have heard my story before about my TNT. It fried ONLY due to the fact that the heatsink on the back was installed incorrectly, it was not a direct result of overclocking. I'm still just a little bit afraid of pushing my V3 to far, but my confidence in it is building. I didn't think I could take it as far as I already did. Oh well. btw someone please post how to get the NFS3 bench.
hey guys, I was just over at FiringSquad.com and they have info on running a Q3 bench in case any of you havn't done it yet. But all you do is 1.) bring down the console 2.) type (timedemo 1) then hit enter then type (demo q3test1) hit enter. 3.) game will load then it will look like it froze looking at a wall but if you give about 2 min it starts to go, yeah I though it was a little weird too but it works. 4.) finally hit the console again and it displays your benchmark
I got 49.3fps on my V3@190mhz in 1024x768!
overclock9 - thanks for the tip on the heatsink. it's been on there for a few days, with no problems, but maybe it's not worth it, with a risk of shorting somthing out.
My other problem - I can't run Q3test!
I downloaded the program, and glsetup, ran both. i can get into the game, and start it up, but all i can see is a black screen with the health and ammo numbers on it. I can jump and run, and i hear the sounds, but i can't get any visuals, (in the actual game)
anyone have any clue what might be wrong? all my other system specs are above. k6-2 400 and V3 3000 64M ram
Buckeye Have you try on www.3dfx.com that have q3 compitable that what it work with my v3 3000 with q3test.
Mine work fine.Hope you make your q3test work after you downla\oad from 3dfx.
Overlock9 I have no idea how to use my fps on q3, can you define??
Aslo i need fps for my nfs3 too.
I let all ya know when i type timedemo 1 on unreal average 38.5 fps on 600x800.
I finally got some time to do a couple of quake3 test benchmarks on my TNT2 Ultra .
heres my system
asus P2B-LS motherboard
256megs of pc100 memory
ibm deskstar 7200rpm harddrive.
creativelabs tnt2 ultra (boxed drivers)
creativelabs sblive (liveware 2.0)
creativelabs 48x cdrom
creativelabs pc dvd 6x with dxr3 card
creativelabs CDRW 4/2/24 speed
sony 3.5 disk drive
2 metabyte wicked voodoo2s in SLI
microstar full tower case with about 7 case fans.
settings: every option in quake3 is set to highest quality visuals , all windows98, creative launcher, and norton system works 2.0 programs running,and no overclocking of cpu or Tnt2.
Processor speed set at 450mhz
TNT2 ultra clock speeds are 150core/183 memory.
16bit color with 16bit texture settings
1024 x 768 = 59.5 fps in timedemo 1
32bit color and 32bit textures settings
1024 x 768 = 49.1fps timedemo 1
I will run some more benchmarks like 32bit color setting with 16bit textures, normal visuals, and so on later.
Rwc, not sure what you mean, did you checkout my post eariler on how to do q3test? Let me know. Hey Tekken, really seems like 16 to 32 bit takes a bigger performance hit in Q3 on UTNT2. I think one of the scores I posted earlier was in 22bit, ours are about the same score in 1024x768. My machine is overclocked too, so from this comparision it looks like UTNT2 should be a better performer for Q3. I bet its due mainly to the fact that the UTNT2 uses True AGP. That's the one thing that I wish my V3 had. Oh and Tarrant its kind of weird to have a better Crusher score than Massive.
[This message has been edited by Overclock9 (edited 06-05-99).]
hey all, i got some q3 benching done, and this is what i got. i'm running v 1.05 of q3test, using high quality sound, light flares disabled (must be default, 'cause i never changed it), running my tnt @115/124, my celery @472, with mobo monitor lite v2.2 running, and in 16 bit.