NVIDIA Blackwell

NVIDIA B200 GPU cloud

Reserve NVIDIA Blackwell B200 capacity on VESSL Cloud: 192GB HBM3e, 8 TB/s bandwidth, and FP4 acceleration for frontier-scale training and real-time inference.

Reserve capacity See pricing & availability

NVIDIA B200

GPU memory: 192GB HBM3e
Memory bandwidth: 8 TB/s

Technical specifications

Architecture: Blackwell
GPU memory: 192GB HBM3e
Memory bandwidth: 8 TB/s
NVLink: 1.8 TB/s
FP8 (Tensor): 9 PFLOPS
FP4 (Tensor): 18 PFLOPS
Max TDP: 1,000W
GPUs per node: 8 (HGX B200)

*Peak performance with sparsity, per NVIDIA official specs. Final specs may vary by node configuration.

Pricing & availability

NVIDIA B200Reserved only

Talk to sales

What's the B200 best for?

Frontier-scale pretraining

FP4 and FP8 throughput plus 1.8 TB/s NVLink make trillion-parameter model pretraining practical in a single Blackwell cluster.

High-throughput inference

192GB HBM3e and FP4 acceleration serve large models at high concurrency.

Multi-modal & generative video

Memory- and compute-heavy workloads (text-to-video, 3D generation, large multi-modal training) fit on a single HGX B200 node.

Compare NVIDIA data-center GPUs

	A100 Ampere	H100 Hopper	H200 Hopper	B200 You're viewing	B300 Blackwell	GB300 Blackwell
Architecture	Ampere	Hopper	Hopper	Blackwell	Blackwell	Blackwell
GPU memory	80GB HBM2e	80GB HBM3	141GB HBM3e	192GB HBM3e	up to 288GB HBM3e	up to 288GB HBM3e
Memory bandwidth	2.0 TB/s	3.35 TB/s	4.8 TB/s	8 TB/s	8 TB/s	8 TB/s
FP8 (Tensor)	Not supported	3,958 TFLOPS	3,958 TFLOPS	9 PFLOPS	9 PFLOPS	10 PFLOPS
Access	from $1.48/hr	from $2.98/hr	Reserved only	Reserved only	Reserved only	Reserved only
Best for	Cost-efficient fine-tuning & inference	Cost-efficient training & inference	Long-context & large-model inference	Frontier-scale training (FP4)	Largest models & reasoning inference	Cluster-scale training & inference

Why AI teams run GPUs on VESSL Cloud

From A100 to Blackwell Ultra

Run A100 and H100 on demand. Reserve H200 and Blackwell GPUs for larger runs, so each workload gets the GPU it needs.

Issues caught early

Every VM Cluster node is health-checked continuously, with its status refreshed about once a minute. Workspaces and Jobs alert you when a run crosses a resource threshold or hits a runtime issue.

From one node to a cluster

Train and serve on up to 8 GPUs on one node, and move to a multi-node VM Cluster when one node is not enough.

Enterprise-ready

SOC 2 Type II compliance, with dedicated support for production AI.

Frequently asked questions

How do I get access to NVIDIA B200 GPUs?

Blackwell capacity is allocated on request. Talk to our team for current availability and pricing.

How much memory does the B200 have?

The B200 offers 192GB HBM3e per GPU at up to 8 TB/s memory bandwidth and 1.8 TB/s NVLink, roughly 2.4× the H100's memory and over 2× the bandwidth.

What's the difference between the B200 and B300?

The B300 (Blackwell Ultra) increases memory to up to 288GB HBM3e (vs 192GB on the B200) and delivers roughly 1.5× the B200's dense FP4 compute, built for the largest models and high-concurrency reasoning.

How does the B200 compare to the H200?

The B200 is a generation ahead: Blackwell architecture, FP4 acceleration, 192GB HBM3e, and 8 TB/s bandwidth vs the Hopper H200's 141GB at 4.8 TB/s. Choose B200 for frontier-scale training and FP4-accelerated inference.

Can I reserve a full B200 cluster?

Yes. We provision HGX B200 nodes (8 GPUs each) joined by high-speed InfiniBand, from a single node to multi-node clusters. How many nodes you can reserve depends on current supply, so talk to our team for availability.