Elsa Gladiac GeForce 2 GTS – ASUS vs SOYO

When I had finished bench marking my new Elsa Gladiac GTS 2 on my Asus P3V4X, I was quite sure I’d swam all I wished to in those waters. All those hours drifting by with only a few numbers to show for it.

Then I began thinking (not always a good sign) what differences would show up in my benchmarks if I ran my Elsa on another motherboard? Say, on my Soyo 7VCA with a 500E at 750 MHz using HSDRAM, as opposed to my Asus with a 650 E at 877 MHz with VCRAM.

I knew that ASUS’s 17% speed advantage would make a head-to-head comparison skew in favor of ASUS. However, we could adjust for it and see if the advantage held through all settings. Also, it would be interesting to see the differences due to HSDRAM vs VCRAM.

Besides a lot of people (I reasoned) probably don’t know whether the Elsa could do well in a Soyo enriched environment. What if someone read my first review of this card and thought that if it did well on a Asus VIA board then it surely must do well upon a Soyo VIA board, and then it didn’t!! I could be sued! I might lose my job!

So as much as I dreaded running all those benchmarks again, on my NEW 3DMARK SOFTWARE I saw I clearly had a duty to see what was what.

Here are the SOYO BENCHMARKS compared to the ASUS. I only felt the need this time to run them once at the default setting. And once at an overclocked setting of 365 MHz Memory clock speed and a GPU core of 225 MHz. After all, the card isn’t that new.

Setting

ASUS Default

ASUS 365/225

SOYO Default

SOYO 365/225

640 X 480 16 bit

7490

7579

7197

7221

1024 X 768 16 bit

6425

6683

6280

6498

1280 X 1024 16 bit

5051

5396

5010

5315

640 X 480 32 bit

6942

7113

6713

6819

1024 X 768 32 bit

4741

5138

4606

4936

1280 x 1024 32 bit

2982

3342

3133

3419

Conclusions

The ASUS speed advantage (17%) did not translate into 17% better scores – the largest difference was +5% (600 x 480 @ 365/225) which diminished as resolution and density increased to a +4.8% advantage for SOYO! Talk about a comeback!

I think what this shows is that CPU speed is important but that other factors, like RAM, can be at least as important, if not more so. I guess the HSDRAM really shines as densities increase, and it may be a better choice for superior video performance.

If you have a video card of your own, which seems likely unless you’re at the library reading this (commendable but don’t you think you can find a better way to utilize valuable library resources?).

Anyway, for those of you that do have a video card of your own, though probably not as NEW as this one, you can download 3DMark and run a mess of benches. Then you can compare your benches to my benches and sulk. Keep on sulking until your wife NOTICES you sulking.

Then you can show her my benches. If you sense resistance, start talking about T&L quad pipelines and stuff until her eyes glaze over. Then slowly walk away mumbling softly “so that’s what I’m going to do OK?”

Later you can say that you explained it all to her and that she agreed. I learned this method of negotiation from my wife. So I can say with certainty it will work on your husband, if you have one of those instead of a wife. I caution you that if you have one of each you probably don’t have time for a new video card

I would like to thank Mike at Hyper Microsystems for sending me benchmarks to show my wife and for vending this card to me.

Email Dan


Be the first to comment

Leave a Reply