Theoretical Performance
51.22 TFLOPS

Graphics Engine
Hopper

Graphics Memory
VRAM, primarily of GDDR type, is a synchronous memory similar to standard RAM. However, graphics memory uses faster throughput and multiple data transfer rates. This results in faster buffering of data calculated by the graphics card or coprocessor and passed to the processor.

Memory Size
80 GB

Memory Type
HBM2e

CUDA Technology
Professional application users benefit from CUDA graphics stream processors, which leverage CUDA architecture. This enables the raw computational power of the graphics card to perform specific calculations, greatly accelerating tasks compared to traditional processors with fewer cores.

Stream Processors
14592

Number of Tensor Cores
456

PCI Express
PCI Express is an interface used for expansion slots, ensuring system modularity with components like GPUs, network cards, controllers, M.2 drives, and more. Newer generations and wider interfaces improve performance and throughput. Modern graphics cards use 16 lanes for processor connections. The latest PCI Express generation 6.0 offers speeds of 7.5 GB/s per lane.

BUS
PCIe 5.0 x16