I don't get it, 9600M GT better than 9650M GT?

According to Notebookcheck.net, nearly every benchmark the 9600M GT beats the 9650M GT. Why is this? I thought the 9650M GT should be better because of the higher clock?

Example: Crysis low settings
9650M GT: 42 FPS avg
9600M GT: 62 FPS avg

 

The cards themselves are the same except for the clockspeed; the 9650M GT has increased clockspeeds over the 9600M GT.

What is more important than clockspeeds is the memory type; a 9600M GT DDR3 will beat a 9650M GT DDR2.

Also, the CPU will come into play in multi-CPU games like Crysis . . . there are several factors that you need to take into account when comparing benchmarks:
-Driver version
-Operating system (64 vs 32 bit makes a difference)
-Hardware (CPU/RAM)

A 9650M GT will be faster than a 9600M GT if it has the same memory type and all of the above factors are constant.

 

You have enough posts here to know notebookcheck is unreliable.

 

Remember there is always the issue with RAM speed, if the 9600M GT is equipped with DDR3 RAM, and the 9650M GT with DDR2, the small clock increase is negated due to much higher DDR3 clock speed. DDR2 vs DDR2 cards, the 9650M GT>9600M GT.

I think I have only seen the 9650GT with DDR2, so yeah, basically 9600M GT DDR3> 9650M GT DDR2

 

ddr3 vs ddr2

well im late...

 

I know that the VRam makes a difference, but benchmarks all around show the 9600M GT is better. The highest 9650M GT is avg for the 9600M GT.

How is notebookcheck.net unreliable? Just because i've been here for a couple of months doesn't mean i read every thread over the past few months.

 

Yeah notebookcheck is a bit unreliable as they test only select versions of some cards. If you look at the ordering of the cards on notebookcheck you can see the 9650M GT is a good half dozen places higher on the list, coming under ati's latest 4670 and the 9700M GT.
i can imagine comparable cards ie same memory type, the 9650M GT would pull ahead on average 500-1000 points

 

Notebookcheck is unreliable in the sense that they're not consistent in their tests and they just average everything.

A good example is the DDR2 vs DDR3. Notebookcheck does not make the distinction and benchmarks both versions of a same GPU then averages the scores. Also note it doesn't take into account different hardware and OS(i.e. it just once again benchmarks everything and averages it).

Arguably, if yo know what you're doing, notebookchek can become a viable source of information, but you have to look at the details of their benchmarking to see specifically what systems scored what, not just look at the overall averages. Also, its of classifications should be taken with a grain of salt as sometimes they list things based on pure on-sheet specs rather than actual performance.