NVIDIA Tesla V100:
NVIDIA® Tesla® V100 is the world’s most advanced data center GPU ever built to accelerate AI, HPC, and Graphics. Powered by the latest GPU architecture, NVIDIA Volta™, Tesla V100 offers the performance of 100 CPUs in a single GPU—enabling data scientists, researchers, and engineers to tackle challenges that were once impossible.
Main Specifications | |
Product Series | Tesla V100 |
Core Type | NVIDIA CUDA |
Host Interface | PCI Express 3.0 x16 |
GPU Architecture | Volta |
Detailed Specifications | |
Streaming Processor Cores | 5120 CUDA Cores |
NVIDIA Tensor Cores | 640 Tensor Cores |
PCIe x16 Interconnect Bandwidth | 32 GB/s |
CoWoS HBM2 Stacked Memory Capacity | 32 GB |
CoWoS HBM2 Stacked Memory Bandwidth | 900 GB/s |
Max Memory Bandwidth | 900 GB/s |
Peak Single Precision floating point performance (GFLOP) | 14 TeraFLOPS |
Peak Double Precision floating point performance (GFLOP) | 7.0 TeraFLOPS |
Tensor Performance | 112 TeraFLOPS |
NVIDIA CUDA™ Technology | Yes |
Cooling | Passive |
Dual Slot | Yes |
Max Graphics Card Power (W) | 250W |