Question for Alienware Reps

Hello,
I'm just curious if you guys have talked to engineering at all about this.
I know the GTX 765m is a GK106 which is a slimmed down GTX 770m.

This leads me to question why the GTX 770m wasn't offered in the AW 14. I mean. If they can solder one version of a GK106 why not a full version?
I suspect it would likely be due to either the extra 5-10W TDP or the extra 64Bit wide bus. I don't really expect an answer however i feel it is a plausible question as i have no doubt it could have handled the heat with the GTX 765M only hitting 75C in extreme conditions.

To people who arent familiar:
GTX 765M ~ 768 Cores 863 Mhz 128Bit GDDR5 1002Mhz
GTX 770M ~ 960 Cores 811Mhz 192Bit GDDR5 1002Mhz

Both based off of GK106.

 

Does the 770m even come in a soldered-on package? That may be why; not enough space to fit an MXM slot.

That's why you'll see the 765m around in smaller laptops, I think.

 

Like i said, the only difference is the 64 Bit memory bus. Its the exact same die with 192 Cores or 1 SMX disabled.


GK106 ~ Manufactured with 5 SMX's availble and trimmed down accordingly

GK104 (GTX 780m & GTX 680m) ~ Manufactured with 8 SMX's available and trimmed down accordingly.

 

Just because it's similar, doesn't mean that nVidia actually offered 770m soldered-on kits as well. I'm not talking about the device at all, I'm just talking about product availability.

 

Yeah, i mean, it sounds like it could be soldered on. Thats basically what i am asking. I don't see how a 192 Bit bus could prohibit that.

 

I understand what you're saying. They most likely could offer a soldered on package for every chip they make, except for maybe the 100w+ beasts. But it unfortunately doesn't mean that they do.

Just to hit on exactly what you're saying, I cannot find another soldered on GPU with a 192bit bus, so maybe that is a limiting factor. Even the 8870m used in Samsung notebooks is 128bit.

Edit: Finally found one. The nVidia 635m was 192bit and soldered on, but it's on the 40nm process. (GF116)

 

I mean it sounds like a win for nvidia, they would be able to move more and they wouldn't be upgrade-able.

 

That's not really true. Look at the 760m as an example. Have you seen in anywhere? I can think of maybe two laptops that are using it. Everyone else is using the 765m as their mid-spec GPU. The reason for that isn't performance, the reason for that is that the higher number dictates a higher price, and more value.

 

Higher cost, higher heat, higher power draw.

The reason they didn't pick the higher end part is because it didn't fall within the design specification allowances for the laptop they were trying to create.



Sent from my DROID RAZR HD using Tapatalk 2

 

Looking to hear if those are the reasons from a Dell employee.

Thanks though.

 

Another factor in this may also be the fact that their competitors only have the 765M, so, they only need the 765M to compete on a level playing field. I'm sure if someone else came out with a 14" laptop that had the 770M, Alienware would have it as well. Another reason may be the possible loss in revenue from the Alienware 17 because it always has two different GPU's offered. The specifications are similar for these two systems, but the 17 has a larger form factor and that allows for a larger price tag. I think if offered, some people who would have originally purchased the 17 may have opted for the 14 because it's cheaper by default.

I'm sure a little of everything mentioned in this thread is the reasoning behind it.