OEM
HGX B300
Blackwell
為 AI 工廠效能樹立全新標準，涵蓋從訓練到推論的完整流程
NVIDIA DGX™ B300 是一款專門設計的 AI 基礎架構解決方案，專為符合生成式 AI 運算需求而量身製作。DGX B300 搭載 NVIDIA Blackwell UltraGPU，提供用於推論的 192 petaFLOPS 以及用於訓練的 70 petaFLOPS，同時採用全新外形規格，可順利融入現代化資料中心。相容於 NVIDIA MGX™機架相容，並搭配完整堆疊軟體，簡化並精簡 AI 部署，讓任何企業都能像超大規模企業般運作。
產品規格
GPU | 8x NVIDIA Blackwell Ultra GPUs |
GPU Memory | 2.3 TB total |
Performance | 72 petaFLOPS FP8 training 144 petaFLOPS FP4 inference |
NVIDIA® NVSwitch™ | 2x |
NVIDIA NVLink Bandwidth | 14.4 TB/s aggregate bandwidth |
System Power Usage | ~14.3kW max |
CPU | 2 Intel® Xeon® Platinum Gen6 Scalable Processors |
System Memory | MAX to 4TB |
Networking | Embedded 8pcs CX8 InfiniBand on GPU – 800G/S networking |
Management Network | 1GbE onboard NIC with RJ45 1GbE RJ45 Host baseboard management controller (BMC) |
Storage | 10 x 2.5” bays (10 x NVMe) |
Rack Units (RU) | 10 RU |
Power Supply | 5+5 redundant 3200W 80 PLUS Titanium PSU |
System Dimensions | 945(L) x 447(W) x 394.5(H) (mm) |
Operating Temperature | 10C° - 35C° |
Enterprise Support | Three-year business-standard hardware and software support |