According to the latest report, Nvidia has taped out its next-generation GP100 Pascal GPU based on TSMC's 16nm FinFET manufacturing process and featuring up to 32GB of second-generation High Bandwidth Memory (HBM2).

While Nvidia might have skipped the first generation of High Bandwidth Memory, which will be used on AMD's upcoming Fiji GPU, it appears that High Bandwidth Memory is the way to go and Nvidia's next-generation Pascal GPU will be using the second-generation High Bandwidth Memory (HBM2).

According to first details coming from Beyond3D Forums, Nvidia new 16nm Pascal GPU, codenamed the GP100, will pack anywhere between 4500 and 6000 CUDA cores, making it the biggest GPU Nvidia ever made. With the new 16nm FinFET manufacturing process and new HBM2 memory, which will offer up to 1200GB/s of memory bandwidth, we will be looking at some really fast graphics cards.

While there are not precise details regarding the GP100 Pascal GPU, Nvidia is apparently targeting Q1 2016 release date.

While GP100 sounds impressive and should bring some new impressive GTX Titan graphics cards, the GP104 GPU should be quite interesting as this one will hit the sweet-spot for consumers and with such impressive specifications for the GP100, the GP104-based GTX 1000 series is something that we certainly can't wait to see.



Source: via Tweaktown.com.