GB300 GPU
The NVIDIA GB300 Grace Blackwell Ultra Superchip connects two B300 GPUs to a Grace CPU via NVLink-C2C, offering 576GB HBM3e memory and massive compute for AI reasoning and trillion-parameter models.
Starting Price
$3.60/hr
Available on 2 cloud providers
Key Specifications
💾Memory
576GB VRAM
🏗️Architecture
Blackwell
⚙️Compute Units
N/A
🧮Tensor Cores
N/A
Technical Specifications
Hardware Details
ManufacturerNVIDIA
ArchitectureBlackwell
CUDA CoresN/A
Tensor CoresN/A
RT CoresN/A
Compute UnitsN/A
GenerationN/A
Memory & Performance
VRAM576GB
Memory InterfaceN/A
Memory Bandwidth16000 GB/s
FP32 Performance200 TFLOPS
FP16 Performance10000 TFLOPS
INT8 PerformanceN/A
Performance
Computing Power
Computational Performance
FP32 (TFLOPS)200
FP16 (TFLOPS)10,000
Common Use Cases
AI reasoning, training trillion-parameter models, large-scale inference, HPC
Machine Learning & AI
- Training large language models and transformers
- Computer vision and image processing
- Deep learning model development
- High-performance inference workloads
Graphics & Compute
- 3D rendering and visualization
- Scientific simulations
- Data center graphics virtualization
- High-performance computing (HPC)
Market Context
The GB300 sits within NVIDIA’s Blackwell architecture lineup, positioned in the ultra performance tier. It’s designed specifically for data center and enterprise use.
Cloud Availability
Available across 2 cloud providers with prices ranging from $3.60/hr. Pricing and availability may vary by region and provider.
Market Position
Released in 2025, this GPU is positioned for enterprise and data center workloads.