Nvidia GP107 GPU will use Samsung’s 14nm technology as opposed to TSMC’s 16nm FinFET process. Furthermore, this means the upcoming GTX 1050 and 1050 Ti could be the first GPUs to use Samsung’s 14nm FinFET manufacturing process. AMD has already been using this on their Polaris GPU. It’s called the GlobalFoundries 14nm FinFET process.
Nvidia GP107 GPU
“The two smallest Pascal chips GP107 & GP108 be noisy (fixed) statements of our forum is not at TSMC in its 16nm production, but at Samsung made in their 14LPP production – which as known represents the technical basis for the 14nm fabrication of GlobalFoundries, in which the Polaris chips are launched by AMD.
This would in any case, completely different clock speeds of the GeForce GTX 1050 also to establish technical. On the other hand, the (considerable) clock rate difference between AMD’s Polaris and Nvidia Pascal should first less by the different production. But rather the different design approach of both graphics chip developer explain. AMD packed very tightly, thus bringing many hardware units on little chip area under what is at the expense of the clock rate. And for Nvidia it is exactly reversed.
As Nvidia this accurately handles the smaller Pascal chips GP107 & GP108, is of course unknown to their release – but unless you want to venture into unfamiliar territory to larger deviations from recent nVidia design approach especially in GP107 & GP108 are more likely than unlikely.
Same principle. Should Nvidia also still meet the supposedly well lower cost point of 14nm production. Because with smaller graphics chips as GP107 and GP108 are the margins known as low. And pays off the whole even in a beachtbaren quantity success. A completely different economic approach than in the larger Pascal graphics chip.”
You can read the full article on 3dcenter.org. Stay tuned t MobiPicker for more news and updates regarding the Nvidia GP107 GPU.