GPU Architecture
NVIDIA Ampere architecture
GPU memory
4 × 16 GB GDDR6
Memory bandwidth
4 × 200 GB/s
Error-correcting code (ECC)
NVIDIA Ampere architecturebased CUDA Cores
4 × 1280
NVIDIA third-generation Tensor Cores
4 × 40
NVIDIA second-generation RT Cores
4 × 10
FP32 | TF32 | TF321 (TFLOPS)
4 × 4.5 | 4 × 9 | 4 × 18
FP16 | FP161 (TFLOPS)
4 × 17.9 | 4 × 35.9
INT8 | INT81 (TOPS)
4 × 35.9 | 4 × 71.8
System interface
PCIe Gen4 (×16)
Max power consumption
250W
Thermal solution
Passive
Form factor
Full height, full length (FHFL) Dual Slot
Power connector
8-pin CPU
Encode/decode engines
4 NVENC/8 NVDEC (includes AV1 decode)
Secure and measured boot with hardware root of trust for GPU
vGPU software support
NVIDIA Virtual PC (vPC)
NVIDIA Virtual Applications (vApps)
NVIDIA RTX Virtual
Workstation (vWS)
NVIDIA AI Enterprise
NVIDIA Virtual Compute Server (vCS)
Graphics APIs DirectX
12.072
Shader Model 5.172
OpenGL 4.683
Vulkan 1.183
Compute APIs
CUDA
DirectCompute
OpenCL™
OpenACC®
MIG support