where do you compare mobile video cards?

That site and ANANDtech have Mobile sections where mobile cards are "benchmarked'. Its a loose guide at best. I would have NEVER EVER considered a 5200Go if I had gone by those articles and charts!!!

I think as a more of a relative measurement of cards to each other they are probably more useful...but as a FPS guide..worthless. I get more FPS than they list for better cards in real world play.

Yeah I don't think you will find that as no one is concerned with the low end cards. You cansee how bashed they are in these forums. These are the guys who support Anandtech and Toms. They don't wasnt to read about how to make a 5200Go run Halflife II well.

I change seetings in a game from the default software/graphics setup. Each game is different on what's important. Thats why I don't think there is a guide for lower end cards. The higher end cards are all tested wide open, everything on (whether its visible or not) Thereofr they can get some sort of baseline number. even then its crap and just numbers a very loose relative "score".

No one says "Well on Max Payne" you should run this setting or that setting, or on COD, This setting is more important..Actually the SOFTWARE says it and adjusts the hardware accordingly. Now we want to Mess with that as if they are WRONG!!! Just like the guys OCing the Radeons because DELL is WRONG in their clock settings! We all think we know more than the people who make what we buy! And they give us the rope to hang ourselves with because we begged to be able to OC and oNLY bought MB's V Cards etc that could be overclocked. So the makers had to make them. Then we wonder why we have freezing , reboots, smoke and like issues!!!

Your question was my question as well BEFORE I bought my machine. I could not get an answer so I bought a low end machine to see for myself based on MY OWN sense of what is good. Don't get wraped up in other's benchmarks or synthetic scores. The best way is to play with it yourself. I downloaded 52.03 drivers and mod INF to do just that. Play and look for differences. Well there isn't much between settings.

Look at screen shots with AA/AF turned on/off. For all the hosepower needed to turn that stuff on, you really have to look to see a difference and you will ultimatly find the experience is just as much fun on either screen,AA/AF on or off.

People buy numbers not facts.....I remember that from the old HiFi buying days at Circiut City. before you could buy computers at a retail store

...It still applies

Your point is well made and well taken. Most benchmarks are artificial and bear only limited relationship to how a particular game will play on a particular machine with a particular GPU and other components and set up in a particular way. Even how someone plays a game can have significant influence.

But there's no way to quantify all that, so they make benchmarks. Unfortunately, with too much emphasis on benchmarks, things like NVidia's tweaking cards specifically to max out the benchmarks at the expense of general usage occur.

It can get you high numbers and generate sales from those who worship benchmarks.

Its like focusing on the specs and deriving the notion that because a centrino machine can match some performance of a standard P4 with twice the clock speed, its a better machine. Clock speed is only one of many interelated issues. The main reason the centrino machines do perform well is their large L2 cache. If you gave a P4 a 1 MB L2 cache, it would smoke. Why they don't, I sure don't know. Its one of those questions like why have they been teaming the Athlon 64 with 64 MB ATI 9600s. Why not give it the full monty? Why, I sure don't know. Perhaps they have a lot of 64 MB ATI 9600s or the cost of a 128 MB would push the price above some marketing limit they believe the machine wouldn't sell well above. Or maybe somebody made up a bunch of mobile mobos that can only accept a 64 MB video system and they want to get rid of them before they do the right thing. What ever it is, its a safe bet there's some financial gain for someone in there somehow. Always is, always will be.

In short, the performance of any machine is the result of a lot of interelated features. Better than a comparison of cards and chips would be a guide on how to tweak them. If you up that, it will get you this, but cost you that. If you drop that, it will cost you this but gain you that. That's priceless information that takes more than just running a benchmark on a couple of different setups. Its also counterproductive to the usual policies and practices of most marketing departments. They want sound bites, quick sexy bits that are easy to convey and get people to buy, buy, buy.

Its like the obsession with FPS, you will notice those that really push higher and higher FPS generally have an interest in selling new graphics systems while those that question excessive FPS generally have nothing to gain. Does that make the arguments of one better or worse than the other, not necessarily but it does give one pause to reflect.

Heat..is the issue. If it were not for heat issues we would have much higher performing machines. Whether or not we NEED or even use the performance we have in today's apps is another story.

Death - I still have 4 full versiosn of that game.. i could never get it to run smooth with an army of air.

Now a days i see little improvement from one video card to another because nealry every card can play smoothly to a degree of how much quality your willing to sacrifice. back in days of voodoo2 where games like quake3 pretty much blew on anything lower than voodoo banshee card regardless of driver or overclocking or reducing of quality.