Half life 2 Article from Max PC

[DemoCoder]

Well then, you better take that up with Anand and other reviews, because I've looked at a few sites, and they all have the MX440 running 100fps at 800x600. Try googling MX440 + UT2003. In fact, even B3D's review scored 100+fps http://www.beyond3d.com/reviews/asus/v9180vs/index.php?p=6

Something is seriously wrong with your gaming rig.

 

[Althornin]

nah, he just isnt thinking flyby.

 

[DemoCoder]

Yeah, but botmatch is CPU limited and we are positing fast CPUs (2+Ghz) paired with value cards, not the other way around.

 

[swaaye]

It was a MX440 AGP 128MB on a P4 2Ghz Northwood with 512MB DDR.

I never ran a flyby benchmark on it. This is just from gameplay experience. If you set some of those options to Highest in UT2003 you will seriously regret it on a GF4MX.

To its credit though a GF4MX seems just as capable as a GF2 GTS+ card. Pretty stunning for a card with half the pipes.

There are people that certainly don't need more than a GF4MX. I think it's a lot better than Intel Extreme 3D (right?). But with the darned OEMs stuffing the things in every stock PC they sell how can we gamers ever begin to see DX8? I can't imagine how many years before DX9 kicks in.

I guess it's just tons of repressed anger being released NVidia should never have labelled NV17 a GF4. I'm sure it made the card look like candy for the OEMs and uber-cheap for NV to make. It's all about making a buck, not advancing the industry. I'll never forgive them for it.

ATI too is scheming a little with their 9000, 9100, and 9200 cards but at least those are DX8. How many times can they contradict themselves on their naming scheme? 9100 9th gen? BS! 9200 DX9? BS!

Uhgg.....

 

[Randell]

Yeah, but botmatch is CPU limited

so is real life gaming. Not synthetic GPU tests.

 

[Gubbi]

Yeah, but botmatch is CPU limited

so is real life gaming. Not synthetic GPU tests.

Only if you play with bots.

Cheers
Gubbi

 

[Kaizer]

Only if you play with bots.

...or any kind of in-game AI, not to mention physics...

Recently, I have noticed that several single player games have AI characters..

With regards
Kjetil

 

[Ilfirin]

Recently, I have noticed that several single player games have AI characters..

I only have a handful of single player games with AI characters. The rest have targets :/

 

[Randell]

Yeah, but botmatch is CPU limited

so is real life gaming. Not synthetic GPU tests.

Only if you play with bots.

Cheers
Gubbi

well apart from the fact that the flybys have no models in them or physics or AI or network overhead.. no no similarity at all between botmatch and real gaming.

Have you actually played UT2003 online with the framerate counter?

Does it resemble your flyby scores or your botmatch scores eh?

 

[DemoCoder]

Yeah, but botmatch is CPU limited

so is real life gaming. Not synthetic GPU tests.

Real life gaming can be either CPU limited or GPU limited based on the design of the game and what resolution you play in. However, if it is CPU limited all the time, than every extra dollar you spent on GPU performance gives far less return. Botmatch makes a Radeon 9700 look 'close' in performance to a GF4200 because botmatch spends so much time burning CPU, that your video card is wasting cycles.

 

[Randell]

Yeah, but botmatch is CPU limited

so is real life gaming. Not synthetic GPU tests.

Real life gaming can be either CPU limited or GPU limited based on the design of the game

Yeah I should have been clearer, I meant in relation to UT2003 tbh.