nVidia RTX PRO 6000 Blackwell WS (24064CudaCores/752TensorCores,96GB GDDR7 ECC,PCI-E16g5,2sloty,4DP,600W), akt.chlazení

Specifications
GPU architecture      NVIDIA Blackwell
NVIDIA® CUDA® Cores      24,064
Tensor Cores     5th Generation
Ray Tracing Cores     4th Generation
AI TOPS      4000 AI TOPS1,2
Single-precision performance     125 TFLOPS1
RT Core performance      380 TFLOPS1
GPU memory     96 GB GDDR7 with ECC
Memory interface     512-bit
Memory bandwidth     1792 GB/s
System interface     PCIe 5.0 x16
Display connectors     4x DisplayPort 2.1b
Max simultaneous displays  4x 4096 x 2160 @ 120 Hz
                                          4x 5120 x 2880 @ 60 Hz
                                          2x 7680 x 4320 @ 60 Hz
Video Engines     4x NVENC (9th Gen)
                           4x NVDEC (6th Gen)
MIG Support     >Up to 4x 24 GB
                        >Up to 2x 48 GB
                        >Up to 1x 96 GB
Power consumption     Total board power: 600 W
Power connector      1x PCIe CEM5 16-pin
Thermal solution     Double-flow-through
Form factor      5.4” H x 12” L, dual slot, extended height
Graphics APIs     DirectX 12, Shader Model 6.6, OpenGL 4.63, Vulkan 1.33
Compute APIs     CUDA 11.6, OpenCL 3.0, DirectCompute