PNY NVIDIA RTX PRO 6000 Blackwell Server Edition (TCSRTXPRO6000SE-PB)

Feature Specification
GPU Architecture NVIDIA Blackwell
CUDA Cores 24,064
Tensor Cores 752 (5th Generation)
RT Cores 188 (4th Generation)
GPU Memory 96 GB GDDR7 with ECC
Memory Interface Width 384-bit
Memory Bandwidth Up to 1,792 GB/s
Display Outputs 4 × DisplayPort 2.1
PCIe Interface PCI Express 5.0 ×16
Maximum Power Consumption 600 W
Cooling Solution Active Cooling (Blower Style)
Form Factor Dual-Slot, Full-Height
Board Dimensions 267 mm (L) × 111 mm (H)
Supported APIs DirectX 12 Ultimate, OpenGL 4.6, Vulkan 1.3, CUDA, OpenCL
VR & AR Support Yes
Display Resolution Support Up to 8K HDR
Error Correction (ECC) Yes (on GDDR7)
NVLink Support No
Target Use Case AI Development, Simulation, Rendering, Scientific Computing