NVIDIA GB10 Grace Blackwell vs NVIDIA GB200 NVL72 GPU 186GB

The GB200 delivers 2.6x the FP32 throughput of the GB10 Grace Blackwell (80 vs 31 TFLOPS dense). The GB200 also carries 58 GB more memory (186 GB vs 128 GB).

GB10 Grace Blackwell: Blackwell, 2025 GB200: Blackwell, 2024
FP32 dense lead
2.6x
GB200: 80 vs 31 TFLOPS
Memory
128 vs 186 GB
58 GB more for the GB200
Bandwidth
273 GB/s vs 8.0 TB/s
GB200 moves data faster
See rental pricing from $10.50/hr Download comparison sheet

Share this page

NVIDIA GB10 Grace Blackwell vs NVIDIA GB200 NVL72 GPU 186GB

A post with this link shows the card on the Image tab as its preview.

Printable version →

Specifications

GB10 Grace Blackwell
GB200
Architecture
BlackwellBlackwell
Launch Year
20252024
Form Factor
IntegratedSXM
Memory
128 GB186 GB
Memory Bandwidth
273 GB/s8.0 TB/s
TDP
180 W—
Process Node
—4nm

Performance (TFLOPS)

GB10 Grace Blackwell
GB200
FP64
No verified data available40 TFLOPS
FP32
31 TFLOPS 80 TFLOPS
TF32
No verified data available1,250 TFLOPS
2,500 TFLOPS sparse
BF16
No verified data available2,500 TFLOPS
5,000 TFLOPS sparse
FP16
No verified data available2,500 TFLOPS
5,000 TFLOPS sparse
FP8
No verified data available5,000 TFLOPS
10,000 TFLOPS sparse
FP6
No verified data available5,000 TFLOPS
10,000 TFLOPS sparse
FP4
500 TFLOPS
1,000 TFLOPS sparse
10,000 TFLOPS
20,000 TFLOPS sparse
INT8
No verified data available5,000 TOPS
10,000 TOPS sparse

FLOPS by Precision

What actually differs

The GB10 Grace Blackwell is the newer part: Blackwell, launched in 2025, against the GB200's Blackwell from 2024. Newer architectures typically add lower-precision formats and better throughput per watt, so check the precision rows your workload actually uses.

Dense throughput for the GB10 Grace Blackwell against the GB200: FP32 31 vs 80 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 128 GB against 186 GB, fed at 273 GB/s versus 8.0 TB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.

For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.

Frequently asked questions

Is the GB10 Grace Blackwell faster than the GB200?

At FP32 precision the GB200 reaches 80 TFLOPS dense against 31 TFLOPS for the GB10 Grace Blackwell. The performance table on this page lists every published precision for both GPUs.

Which has more memory, the GB10 Grace Blackwell or the GB200?

The GB200 carries 186 GB of memory versus 128 GB for the GB10 Grace Blackwell. Memory bandwidth is 273 GB/s for the GB10 Grace Blackwell and 8.0 TB/s for the GB200.

Can I rent the GB10 Grace Blackwell or the GB200 in the cloud?

Yes. Live cloud listings tracked by Flopper start at $10.50 per GPU hour. The rental pricing section on this page lists current providers and rates for both GPUs.

Where to Rent

NVIDIA GB200 NVL72 GPU 186GB

ProviderConfigurationPrice/GPU-hrChecked
4× GB200 NVL72 GPU $10.50 3h ago View →
1× GB200 NVL72 GPU $16.00 3h ago View →
All GB200 listings and price history →

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.

© 2026 Flopper.io - Compare the GPUs Powering AI