Nvidia Pascal GP100 is a Powerhouse with 4-Hi HBM2 & 4096-bit Memory Bus
We reported back in June that Nvidia’s new Pascal GPU was tapped out and in final stages but now we are way ahead of that. More details have been shared for the Nvidia Pascal GP100.
What is expected to be the greatest of their products, this baby is going to make use of a gigantic 4096-bit memory bus.
While the consumer standard GP100 GPU is going to have 4-Hi stacks as compared to the industrial Quadro and TESLA GPUs that are getting double i.e. 8-Hi stacks.
Of course that comes together with promised improvements over HBM to HBM2 that is also going to offer other capacity and performance improvements.
You can also expect Nvidia Pascal GP100 to achieve magnanimous throughput because the consumer variant is going to be running on 4096-bit memory bus with HBM2 on 1GHz.
Then there is the TSMC 16 nano meter FinFET
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.
Last but not the least, Nvidia Pascal GP100 is going too bring you five to 12 times faster transfer speeds thanks to the recently developed NVLink technology of Nvidia.
So yeah you have all the reasons to rage on in favor of it.