Improve video without an upgrade?

Hey Goofy,

I notice no mention of a CPU fan. :rolleyes:

If you are locking due to heat: CPU, GPU, RAM &/or chipset. When it locks, can you touch the vid card? Is it *hot*? How 'bout the others? Need to pinpoint what is getting hot (if it is) & then proceed from there.

I just noticed you have a VIA chipset. Do you have the latest 4in1 drivers? Very important, I hear. If the chipset is getting hot, you can attach a heatsink to it (provided there is room) & if it is the CPU: ORB time possibly. If you don't have a CPU fan & only a hs on it: you really do need a fan on it now that the card/games are demanding more out of it then the 3dBlade demanded, IMO.

I've had 3DMark lock (lots), GLExcess (a few) but only once out of literally 100's of runs of FD. :blush: I do seem to have the best of luck w/these programs tho'.

Does Gigabyte offer a utility to monitor temps? That would be one way to check the 'heat' issue. This may work on that board: http://mbm.livewiredev.com/

C'ya 'round,

Yes, there is a CPU fan. I'll let you know how things progress after I try out a few of your ideas.

Later!

Goofer:grinthumb

I had to re-read the thread to remember if you went AGP or PCI card. I'm wondering if the VIA AGP functions are still enabled & are dragging the system down. Usually just installing the card activates an automatic shutdown of the on-board graphics, but ...

Arris had an interesting post on his Gigabyte experience & his system locking too. I've read on other sites the VIA chips are problematic.

Here's a source for excellent BIOS info. I'm not up on the details of BIOS adjustments & will be reading the info there too. Have you installed the latest AMI Award BIOS for that board?

Here's the BIOS explained & I recommend it to anyone that reads this: http://www.itcp.net/~awsh/Speed_Demonz/BIOS_Guide/BIOS_Guide_Index.htm It's very long & highly detailed, as it should be! :grinthumb

C'ya l8tr,

PS: thought you did have a CPU fan, but had to ask.

Conductive paste

Some chips come with a hs, but then they forget to properly connect it. I don't know how yours is connected, but on another thread there was talk of a conductive glue/paste for hs that let more heat transfer. Anybody have an opinion on this stuff?

Re: benchmark

Since then, I have a new monitor, 17" flat-screen. Here are the benchmark results with the new monitor(and it didn't crash).

Display Adapter Benchmark
This module performs display adapter benchmark in your system.



Benchmark Result
Random Pixels 466 KPixels/s
Lines 18,000 Lines/s
Circles 13,341 Circles/s
Rectangles 38,790 Rectangles/s
Texts 517,008 Chars/s
FloodFills 547,840 KPixels/s
Draw 33,961 KPixels/s

Benchmark Info
Device Name NVIDIA RIVA TNT2 Model 64/Model 64 Pro
Device ID \\.\DISPLAY1
Benchmark Mode 800 x 600 pixels
Color Depth 16 Bit

Strange, though, that my Lines and Floodfills are down. The Lines #'s are down significantly. Any reasons why?

Hmmm. :rolleyes:

I tried figuring out FD & even wrote FD for some help in deciphering the scores. Sad to say they haven't replied & I have been unable to determine defineative 'tweeks' that affect the scores 'across the board'.

My own 'experiments' concluded that raising the core clock on the card affected the random pixels, lines & flood fills upwards. Memory ocing affected the circles, rect. & draw. Both had limits & going too high actually brought #'s down.

I would not have imagined a monitor change would have affected your card as it did. Your 'random' increasing where it is > wow! especially when the lines dropped & didn't go up with it. Now I'm wondering about MY 3yr old monitor. :blush:

I'm stumped, Goofy. I can't imagine the refresh rate of the monitor would have the affects we're seeing & if it does, why? What refresh rate did you run & what are you running now?

I'm gonna try different refresh rates & see if it affects FD similiarly here.

Do you have FRAPS? It is a frame rate counter: www.fraps.com I should have mentioned it before & we'd have actual STBC #'s to compare against. Have you noticed any improvements in STBC since the monitor?

HTH,

Jav:

First of all, I am unsure of the monitor refresh rates. How do you determine them?

As for STBC, it works awesome! It gets just a little slow when I am battling 15 Romulan dogs in Battle Mode, because of the extra graphics and AI needed to run them, but otherwise it has been flawless. I am thinking of upgrading the video again to a GF2 after I get working for a while(just finishing University).
I'll give the FRAPS a go. Don't forget to let me know how the refresh stuff is done.

Goof

Goof,

Go into control panel > display > settings > advanced > adapter & there is a box to change refresh rate.

Those Warbirds are tough! Birds of Prey go down easier. 15 ships (including mine) is the max I can run w/o staggers & depending on the intensity of the 'space graphics' even that stutters in heavy combat. I'm a Celeron 500 tho'.

I'll try the different refresh rates tonight & let you know what FD changes (if any) they make on my system.

Jav:

The fefresh rate is 60Hz. Is that the industry standard? Does it need to be changed? Is it safe to change it? I found the birds of prey pretty soft once you gave them the 'what for' with a few quantum torpedoes. The warbirds are hard as granite, and those Kessok buggers are nasty. It's a great game, though. I hope they come out with more mission disks for it.

Goofy

Well I ran FD @ 60, 70, 72, 75 & 85mhz RR & really didn't notice any significant differences.

I didn't try 'optimal' or 'adapter default' because my monitor won't support over 85mhz @ 1024x768 & I don't know what optimal/default is.

I run 85mhz 24/7 w/o adverse affects. You should check your monitor specs (manual) for supported mhz @ the different resolutions (6x8, 10x7, 12x9, etc) & run what works best for you.

I think you're at the 'tweek & test' stage for your comp. Tweek it, start Fraps, open STBC & set-up a battle & watch the fps. Change the tweek & do it again w/the same battle. Trial & error 'til you achieve high fps & stability.

I will suggest you close all programs not needed (anti-virus, firewall, etc.) when you do the tests. :grinthumb

Have you run 3D Mark 2001 or GLExcess benchmarks? They are big dl's (do it @ night when you're sleeping), but they are helpful to see the results (fps wise) of your tweeking. Guy at 3D has a 800PIII & ran 259 w/a TNT card. :dead: I'm thinking you can break 1000 easy.

HTH,

JAV
GBA!

now using different video card

Benchmark Result
Random Pixels 300 KPixels/s
Lines 19,666 Lines/s
Circles 12,140 Circles/s
Rectangles 48,252 Rectangles/s
Texts 476,279 Chars/s
FloodFills 1,569,760 KPixels/s
Draw 27,750 KPixels/s

Benchmark Info
Device Name NVIDIA GeForce DDR
Device ID \\.\DISPLAY1
Benchmark Mode 800 x 600 pixels
Color Depth 16 Bit

This is a 32 mb ddr card using agp. the results are not a great deal more that the old pci card, except for the flood fills.

Benchmark Result
Random Pixels 366 KPixels/s
Lines 219,333 Lines/s
Circles 13,115 Circles/s
Rectangles 65,058 Rectangles/s
Texts 508,456 Chars/s
FloodFills 9,837,760 KPixels/s
Draw 27,852 KPixels/s

Benchmark Info
Device Name NVIDIA GeForce DDR
Device ID \\.\DISPLAY1
Benchmark Mode 800 x 600 pixels
Color Depth 16 Bit


Last Benchmark Friday, Feb 7, 2003 01:38 pm


No changes made to system. Why are the results so different from the last benchmark?

Display Adapter Benchmark
This module performs display adapter benchmark in your system.



Benchmark Result
Random Pixels 933 KPixels/s
Lines 123,333 Lines/s
Circles 28,630 Circles/s
Rectangles 142,848 Rectangles/s
Texts 1,049,152 Chars/s
FloodFills 5,512,480 KPixels/s
Draw 47,171 KPixels/s

Benchmark Info
Device Name NVIDIA GeForce4 Ti 4600
Device ID \\.\DISPLAY1
Benchmark Mode 800 x 600 pixels
Color Depth 16 Bit

Operating System System Model
Windows XP Professional (build 2600) No details available
Processor a Main Circuit Board b
2400 megahertz Intel Pentium 4
8 kilobyte primary memory cache
512 kilobyte secondary memory cache Board: Gigabyte Technology Co., Ltd. 8PE667U
Bus Clock: 133 megahertz
BIOS: Award Software International, Inc. 6.00 PG 10/16/2002
Drives Memory Modules c
202.97 Gigabytes Usable Hard Drive Capacity
66.08 Gigabytes Hard Drive Free Space

DVD-ROM [CD-ROM drive]
HL-DT-ST CD-RW GCE-8400B [CD-ROM drive]
Image Simulation SCSI CdRom Device [CD-ROM drive]
3.5" format removeable media [Floppy drive]

MAXTOR 6L080J4 (80.05 GB) [Hard drive] -- drive 2
Maxtor 6Y060L0 (61.49 GB) [Hard drive] -- drive 1
Maxtor 6Y060L0 (61.49 GB) [Hard drive] -- drive 0 512 Megabytes Installed Memory

Slot 'A0' has 512 MB
Slot 'A1' is Empty
Local Drive Volumes

c: (on drive 0) 30.73 GB 21.87 GB free
d: (on drive 1) 61.48 GB 37.11 GB free
e: (on drive 2) 80.03 GB 3.00 GB free
f: (on drive 0) 30.73 GB 4.09 GB free



My spec which I achieved that score/result if you're interested.

Seeing you have a 2.4 P4 and I am running a Celeron 1.0, I am not surprised. What kinf of vid card are you using? I scored 1960 on 3D Mark 2001.

it wasn't a boast, but other people are posting for comparisons etc, it's a GF4 Ti4600 as it says in the post previous

Yes, I appreciate the ability to compare with higher-end systems. I am pleased with my results now seeing what I am running. (see my profile to see my system specs).

Did you try the 3D Mark benchmark? Awesome! It is very entertaining as well as educating.

New 3D Mark 2001 score of 2014! Will disable all background apps and try again.

** Edit: With latest Detonator drivers, 3D Mark 2001 SE score of 2760! Not bad, seeing my first score was below 2000.

are you running 3dmark2001 se using the default settings?

I haven't changed any settings on the benchmark. After loading the newest Detonator drivers, I maxxed out at 3200. Not bad for this system. Do people usually change the settings on 3d Mark 2001 SE?

yes, actually some people do to get better scores. i am wondering how you get such good scores? i am getting 2441 and i have a
1300mhz duron@ 1400mhz
64mb geforce4 mx440
256mb ddr 2100

.............maybe someone can tell me what the deal is






wait wait, i just checked your profile and you have alot of ram, that helps alot in 3dmark