What's the difference between the GT 650M in the M14x R2 and the GTX 660M?

Perhaps I'm confused but notebookcheck lists the GTX 660M 384@835MHz with GDDR5.

My M14x R2 has 384 Shaders and runs at exactly 835MHz with GDDR5.


The GeForce GT 650M has 384 shaders also but runs at 735 - 850MHz with either memory.


But then if the M14x R2 "has" a GTX 660M then why don't they market it as a GTX 660M?

 

We just have a GT 650m GDDR5. The cards have differences. The 660m is considered "large" and has a much higher memory speed. We don't have a GTX 660M so they'd be lying if they marked it as such.

 

@set12 youre absolutely right. the gtx 660m and the gt 650m are the same kepler cards but with the difference of qualification.

Thats also the reason why the gt 650m`s are performing so extreme different when it comes to OCing.

So gtx 660m`s that dont qualify by some of the nvidia tests they lower the tdp and boost by 100mhz and declare them as gt 650m.

I for example got a very good gt 650m (mid good gtx 660m) that i can stable OC to a higher performance than a standard clocked gtx 660m with a special VBIOS mod.

Alot of gt 650m owners have a gt 650m`s which was performing bad as a gtx 660m at higher core clock and more tdp so they cant OC it too near to a gtx 660m`s performance.

Its the same as with CPU´s. All are produced as 3820qm`s but its the qualification in the tests that degrades them to 3610qm´s or 3720qm´s. Or the performance in the tests was fully flawless and they get the name 3920xm and open multipliers +10tdp.

 

edit: The reason for this behaviour is simply that the production costs of chips is too high to sort faulty or partially faulty chips out.

On the other hand its good for us cause we get these 2nd or 3rd class chips much cheaper than the flawless or very good chips.

Otherwise check the prices if a M14x R2 would only have the choice of having a 3820qm or even 3920xm with a gtx 660m to configure. I bet only very few people could spend such amounts of money.