We’ve haven’t heard that much about NVIDIA’s Pascal GPU since it was taped out back in June. According to many sources the Pascal GPU will be manufactured on TSMC’s 16nm FinFET node. This means that Samsung has failed to win the contract for the latest GPU. The NVIDIA Pascal GPU will be the latest graphics architecture from the company and will feature HBM2, NVLINK, and Mixed Precision Support.
We are not really sure why NVIDIA decided to go with TSMC over Samsung, there were early reports that NVIDIA would use both semiconductor companies to mass produce Pascal GPUs, but it looks like they are going solely with TSMC. It is interesting as Samsung already has 14nm FinFET in production as demonstrated by Apple with their A9 SoC.
Since the GP100 has already been taped out we could see graphics cards based on the chip as early as Q2 2016. The GP100 will have almost double the transistor count as the current flagship GM200 CPU core, so that would put it anywhere between 16-17 billion!
TSMC’s 16FF+ (FinFET Plus) technology can provide above 65 percent higher speed, around 2 times the density, or 70 percent less power than its 28HPM technology. Comparing with 20SoC technology, 16FF+ provides extra 40% higher speed and 60% power saving. By leveraging the experience of 20SoC technology, TSMC 16FF+ shares the same metal backend process in order to quickly improve yield and demonstrate process maturity for time-to-market value.