Just last week NVIDIA announced the brand new Titan X Pascal-based graphics card. At the time it was the fastest consumer graphics card offering which targeted both gamers and PC enthusiasts. Well at SIGGRAPH 2016 NVIDIA has just announced the new Quadro P6000 as well as the P5000. The Quadro P6000 is based on the same GP102 silicon as the Titan X, but the P6000 maxes out all resources on the chip. The P5000 on the other hard is based off the GP104 silicon.
The NVIDIA Quadro P6000 has all 3940 CUDA cores physically present on the chip and features a maximum FP32 (single-precision floating point) performance of up to 12 TFLOP/s. The card will also come with 24 GB of GDDR5X memory, spread across a 384-bit wide memory interface.
The Quadro P5000 is based on the GP104 silicon as features 2560 CUDA cores with up to 8.9 TFLOP/s FP32 performance. It will also have 16 GB of GDDR5X memory across a 256-bit wide memory interface.
Neither cards will feature full FP64 (double-precision) technology, as that is reserved for NVIDIA’s Tesla P-Series.
Both cards will be available in October 2016.