Nvidia drops Samsung and uses TSMC for Pascal

So we already mentioned would very likely be made on TSMC 16nm FiNFET. The target release date for this GPU is 2016. The product would be released as GP100 and will be the succesor to the GM200 series GPUs. As it now turns out Samsung indeed was dropped.

Previously, Samsung Electronics competed with TSMC to win a contract to produce the Pascal GPU. According to industry sources on Sept. 15, Nvidia decided to let TSMC mass produce the Pascal GPU, which is scheduled to be released next year, using the production process of 16-nm FinFETs. Some in the industry predicted that both Samsung and TSMC would mass produce the Pascal GPU, but the U.S. firm chose only the Taiwanese firm in the end. Since the two foundries have different manufacturing process of 16-nm FinFETs, the U.S. tech company selected the world's largest foundry for product consistency.

Samsung has strengthened its competitiveness using the production process of 14-nm FinFETs, which it introduced before TSMC did. In particular, the fact that the A9 processors that are featured in the iPhone 6s were produced using the Korean company's production process of 14-nm FinFETs worried the Taiwanese firm's executives. However, the selection of TSMC, which has been Nvidia's partner company for 20 years, is a setback for Samsung, which anticipated a dramatic turnaround in the foundry market.

The reason for Samsung's determination to win the contract for the Pascal GPU lies in the fact that Nvidia's new GPU is highly likely to mark a milestone in the next-gen graphic market.

Experts are saying that Samsung's failure to obtain the contract is mainly attributable to its lack of experience. The fact that the Korean tech giant has become TSMC's rival only two years after it started to produce GPUs itself is considered to have special meaning at the moment.

- See more at: http://www.businesskorea.co.kr/article/ict/12092/no-samsung-gpu-samsung-electronics-fails-win-contract-nvidias-pascal-gpu#sthash.nDSW53sc.dpuf

With the pascal architecture Nvidia as well will make a move towards HBM memory (stacked on die memory), however they will jump right towards generation v2 of HBM, meaning they could inject up-to 32GB pf memory on the GPU.  The new details show that the flagship NVIDIA Pascal (Single chip) would feature up to 16 GB HBM2 VRAM that can do 1 TB/s bandwidth. 

It is estimated that ther shader processor count for big pascal is anything from 4500 towards 6000 units. Big Pascal made on TSMC's 16 nanometer silicon fab process with a release in Q1 2016.