GPU를 통한 데이터센터 가속화
SYSTEM SPECIFICATIONS (PEAK PERFORMANCE) |
NVIDIA A100 for NVIDIA HGX™ |
NVIDIA A100 for PCIe |
---|---|---|
GPU Architecture | NVIDIA Ampere | |
Double - Precision Performance | FP64: 9.7 TFLOPS FP64 Tensor Core: 19.5 TFLOPS |
|
Single - Precision Performance | FP32: 19.5 TFLOPS Tensor Float 32 (TF32): 156 TFLOPS | 312 TFLOPS* |
|
Half - Precision Performance | 312 TFLOPS | 624 TFLOPS* | |
Bfloat16 | 312 TFLOPS | 624 TFLOPS* | |
Integer Performance | INT8: 624 TOPS | 1.248 TOPS* INT4: 1,248 TOPS | 2,496 TOPS* |
|
GPU Memory | 40 GB HBM2 | |
Memory Bandwidth | 1.6 TB/sec | |
Error - Correcting Code | Yes | |
Interconnect Interface | PCle Gen4: 64 GB/sec Third generation NVIDIA NVLink: 600 GB/sec** |
|
Form Factor | 4/8 SXM GPUs in NVIDIA HGX™ A100 | PCIe |
Multi - Instance GPU (MIG) | Up to 7GPU instances | |
Max Power Consumption | 400W | 250W |
Delivered Performance for Top Apps | 100% | 90% |
Thermal Solution | Passive | |
Compute APIs | CUDA, DirectCompute, OpenCL™, OpenACC® |
* Structural sparsity enabled
* SXM GPUs via HGX A100 server boards; PCIe GPUs via NVLink Bridge for up to 2 GPUs
SPECIFICATIONS | V100 PCle | V100 SXM2 | V100S PCle |
---|---|---|---|
GPU Architecture | NVIDIA Volta | ||
NVIDIA Tensor Cores | 640 | ||
NVIDIA CUDA® Cores | 5,120 | ||
Double - Precision Performance | 7 TFLOPS | 7.8 TFLOPS | 8.2 TFLOPS |
Single - Precision Performance | 14 TFLOPS | 15.7 TFLOPS | 16.4 TFLOPS |
Tensor Performance | 112 TFLOPS | 125 TFLOPS | 130 TFLOPS |
GPU Memory | 32 GB/16 GB HBM2 | 32 GB/1 6GB HBM2 | |
Memory Bandwidth | 900 GB/sec | 1134 GB/sec | |
ECC | Yes | ||
Interconnect Bandwidth | 32 GB/sec | 300 GB/sec | 32 GB/sec |
System Interface | PCle Gen3 | NVIDIA NVLink™ | PCle Gen3 |
Form Factor | PCle Full Height/Length | SXM2 | PCle Full Height/Length |
Max Power Comsumption | 250W | 300W | 250W |
Thermal Solution | Passive | ||
Compute APIs | CUDA, DirectCompute, OpenCL™, OpenACC® |
SPECIFICATIONS | |
---|---|
GPU Architecture | NVIDIA Turing |
NVIDIA Turing Tensor Cores | 320 |
NVIDIA CUDA® Cores | 2,560 |
SINGLE PRECISION PERFORMANCE (FP32) | 8.1 TFLOPS |
MIXED PRECISION (FP16/FP32) | 65 TFLOPS |
INT8 정밀도 | 130 TOPS |
INT4 정밀도 | 260 TOPS |
GPU Memory | 16 GB GDDR6 300 GB/sec |
ECC | Yes |
Interconnect Bandwidth | 32 GB/sec |
System Interface | x16 PCle Gen3 |
Form Factor | Low-Profile PCIe |
Thermal Solution | Passive |
Compute APIs | CUDA, NVIDIA TensorRT™, ONNX |
제품 자체가 AI 데이터센터인 슈퍼컴퓨터
SPECIFICATIONS | ||
---|---|---|
NVIDIA DGX Station A100 320GB | NVIDIA DGX Station A100 160GB | |
GPUs | 4x NVIDIA A100 80 GB GPUs | 4x NVIDIA A100 40 GB GPUs |
GPU Memory | 320 GB total | 160 GB total |
Performance | 2.5 petaFLOPS AI 5 petaOPS INT8 |
|
System Power Usage | 1.5 kW at 100-120 Vac | |
CPU | Single AMD 7742, 64 cores, 2.25 GHz (base)-3.4 GHz (max boost) | |
System Memory | 512 GB DDR4 | |
Networking | Dual-port 10Gbase-T Ethernet LAN Single-port 1Gbase-T Ethernet BMC management port |
|
Storage | OS: 1x 1.92 TB NVME drive Internal storage: 7.68 TB U.2 NVME drive |
|
DGX Display Adapter | 4 GB GPU memory, 4x Mini DisplayPort | |
System Acoustics | < 37 dB | |
Software | Ubuntu Linux OS | |
System Weight | 91.0 lbs (43.1 kgs) | |
Packaged System Weight | 127.7 lbs (57.93 kgs) | |
System Dimensions | Height: 25.1 in (639mm) Width: 10.1 in (256mm) Length: 20.4 in (518 mm) |
|
Operating Temperature Range | 5°C to 35 °C (41°F to 95 °F) |
NVIDIA A100 기반의 세계 최초 AI 시스템
SPECIFICATIONS | ||
---|---|---|
NVIDIA DGX A100 640GB | NVIDIA DGX A100 320GB | |
GPUs | 8x NVIDIA A100 80 GB GPUs | 8x NVIDIA A100 40 GB GPUs |
GPU Memory | 640 GB total | 320 GB total |
Performance | 5 petaFLOPS AI 10 petaOPS INT8 |
|
NVIDIA NVSwitches | 6 | |
System Power Usage | 6.5 kW max | |
CPU | Dual AMD Rome 7742, 128 cores total, 2.25 GHz (base)-3.4 GHz (max boost) | |
System Memory | 2 TB | 1 TB |
Networking | 8x Single-Port Mellanox ConnectX-6 VPI 200Gb/s HDR InfiniBand 2x Dual-Port Mellanox ConnectX-6 VPI 200Gb/s Ethernet | 8x Single-Port Mellanox ConnectX-6 VPI 200Gb/s HDR InfiniBand 1x Dual-Port Mellanox ConnectX-6 VPI 200Gb/s Ethernet |
Storage | OS: 2x 1.92TB M.2 NVME drive Internal Storage: 30TB (8x 3.84TB) U.2 NVME drives | OS: 2x 1.92TB M.2 NVME drives Internal Storage: 15TB (4x 3.84TB) U.2 NVME drives |
Software | Ubuntu Linux OS Also supports: Red Hat Enterprise Linux CentOS |
|
System Weight | 271.5 lbs (123.16 kgs) max | |
Packaged System Weight | 359.7 lbs (163.16 kgs) max | |
System Dimensions | Height: 10.4 in (264.0 mm) Width: 19.0 in (482.3 mm) max Length: 35.3 in (897.1 mm) max |
|
Operating Temperature Range | 5°C to 30 °C (41°F to 86 °F) |
A Global Leader of AI Appliance