NVIDIA GH200 Grace Hopper
14,592
CUDA Cores
96GB
VRAM
3000
GB/s
Technical Specifications
14,592
CUDA Cores
1410
Base MHz
1830
Boost MHz
96GB HBM3
5120-bit bus
Performance
67
FP32 TFLOPS
2000
FP16 TFLOPS
1000W
TDP
Cloud Availability
1
Available Instances
$1.99/hr
Starting Price
Detailed Specifications
| Architecture | Hopper (Unknown) |
| Release Date | 2023-05-29 |
| Launch Price | $50,000.00 |
| Process | 4nm |
| Transistors | 80B |
AI Features
Gen 4
Tensor Cores
Enabled
Transformer Engine
Not Supported
Flash Attention
Physical Specifications
Dimensions
12in
Length
5in
Width
3-slot
Height
About GH200 Grace Hopper GPU
The NVIDIA GH200 Grace Hopper was introduced in 2023 on the Hopper architecture. Its specification combines 14,592 CUDA cores with 96GB of HBM3, which defines both the workloads it can run and the batch sizes it can hold.
Based on its 96GB memory capacity, the GH200 Grace Hopper is best evaluated for large-model training, high-batch inference, and memory-intensive scientific computing. Always size the model, optimizer state, KV cache, and framework overhead together rather than choosing a GPU from core count alone.
GPUvec currently tracks 1 cloud listing for the GH200 Grace Hopper, with the lowest observed hourly rate at $1.99. Prices and availability can change by provider, region, and instance configuration, so verify the final rate before starting a workload.
For software planning, the GH200 Grace Hopper combines 3,000 GB/s of memory bandwidth, CUDA compute capability the latest, and a 4nm manufacturing process. Confirm the minimum CUDA, PyTorch, TensorFlow, or driver version required by your application before renting or purchasing hardware.
Rent GH200 Grace Hopper from Our Partners
Get started quickly with these trusted GPU cloud providers. We may earn a commission when you sign up.
External Resources
Learn more about GPUs from these authoritative sources:
Official CUDA programming guide
NVIDIA GPU Specifications →Official NVIDIA GPU specs
TechPowerUp GPU Database →Comprehensive GPU specifications
CUDA Compute Capability Guide →GPU compute capability reference
What You Need to Know About the GH200 Grace Hopper
Complete Specifications for the NVIDIA GH200 Grace Hopper
The NVIDIA GH200 Grace Hopper uses the Hopper architecture and combines 14,592 CUDA cores with 96GB of HBM3. GPUvec records 3,000 GB/s memory bandwidth alongside architecture, release date, process node, accelerator features, and physical specifications so the page can be evaluated from source data instead of model-name assumptions.
Compare NVIDIA GH200 Grace Hopper Cloud Rental Prices per Hour
GPUvec currently tracks 1 cloud listing, with an observed starting rate of $1.99 per hour. Provider rates can vary by region, host configuration, billing model, and availability. Use the provider directory and related-GPU links to compare an available alternative when this exact accelerator is not listed.
Is the NVIDIA GH200 Grace Hopper the Right GPU for Your AI Workload?
Start with the 96GB memory limit, then evaluate 3,000 GB/s memory bandwidth, compute features, and software compatibility for your workload. Model weights, KV cache, activations, optimizer state, and framework overhead all consume memory. Compare this page with related accelerators rather than assuming that a higher core count alone guarantees a better training or inference result.
Top GPUs for Training and Inference
| Category | Rank 1 | Rank 2 | Rank 3 |
|---|---|---|---|
| Best for Training | NVIDIA H200 | NVIDIA H100 | NVIDIA B200 |
| Best for Inference | NVIDIA A40 | NVIDIA A100 | NVIDIA A10 |
Compare GPU specifications and cloud instances to find the best GPU for your workload.