NVIDIA has just announced the Quadro M6000, which is their high-end workstation single-GPU graphics card solution. It is based off the Maxwell GM200 silicon and maxes out the hardware features with 3072 CUDA cores, 192 TMUs, 96 ROPs, and a 384-bit wide GDDR5 memory interface. The card will hold 24 GB of GDDR5 memory, which is double that of the GeForce GTX Titan X. The card has a peak single-precision floating point performance of 7 TFLOPS/s.
Of course the Quadro M6000 is made for the workstation so that is where it differs from the Titan X as it has workstation-grade features. Instead of the HDMI 2.0 connector you have four DisplayPort 1.2 connectors, supporting a total of four 4K Ultra HD displays. NVIDIA does keep the dual-link DVI connector though. There is also a stereoscopic 3D connector. The nView MultiDisplay tech provides more flexible display-head configurations than the ones you find on NVIDIA’s consumer graphics GPUs. There is also NVIDIA GPUDirect support, which gives you better memory sharing access for multi-GPU systems. As far as API support you have all the modern APIs like DirectX 12, OpenGL 4.5, and Vulkan as well as compute capabilities over CUDA, OpenCL, and DirectCompute. No word yet on pricing.