NVIDIA GB300 NVL72
Grace Blackwell Ultra Rack-Scale GPU
72x B300 · 36x Grace CPU · 20.7 TB HBM3e · 130 TB/s NVLink · 1.4 EFLOPS FP4 · 1024+ GPUs available
The NVIDIA GB300 NVL72 is the Blackwell Ultra rack-scale system pairing 72 B300 GPUs with 36 Grace ARM CPUs over NVLink 5 at 130 TB/s of aggregate bandwidth. Each B300 ships with 288 GB of HBM3e (50% more than B200) and 8 TB/s of memory bandwidth, giving each NVL72 rack 20.7 TB of unified HBM and 1.4 EFLOPS of FP4 dense compute. It is the highest-VRAM rack-scale GPU system NVIDIA produces, purpose-built for frontier-scale pre-training, 200B+ inference serving, and trillion-parameter MoE workloads. Spheron rents GB300 NVL72 across data center partner regions in any count, from a single 8-GPU node to multi-rack deployments.
Tell us how many GPUs you need and what you're running. Our team confirms availability within one business day and gets you provisioned.
Submit Your Request
Tell us how many GPUs you need. Our team confirms availability within one business day and gets you provisioned.
Rent NVIDIA GB300 GPUs on Spheron. 1024+ GPUs are live across our data center partner regions and ready to provision. Rent any quantity, from a single 8-GPU node to multi-rack deployments. Each GPU is a Blackwell Ultra B300 with 288 GB HBM3e. Inside an NVL72 rack, 72 GPUs and 36 Grace CPUs share a single NVLink 5 domain at 130 TB/s, exposing 20.7 TB of unified memory for trillion-parameter training and frontier inference. Submit the form with your GPU count and our team confirms availability within one business day. For workloads that don't need GB300 specifically, B300 per-GPU rentals and B200 are on per-minute billing today.
Where GB300 sits in the stack
GB300 is NVIDIA's latest Blackwell Ultra GPU. Spheron has 1024+ available to rent at any quantity. Rubin R100 is the next generation, available H2 2026.
GB300 NVL72 specifications
Per-rack specifications. Smaller rentals inherit the same architecture at the slice they're allocated. All systems run in NVL72 reference configuration with liquid cooling, 2:1 InfiniBand fat-tree fabric across racks, and persistent NVMe per chassis.
GB300 NVL72 vs GB200 NVL72 vs HGX H100
| Spec (per rack) | GB300 NVL72New | GB200 NVL72 | HGX H100 (8x) |
|---|---|---|---|
| Architecture | Blackwell Ultra | Blackwell | Hopper |
| GPUs / rack | 72 × B300 | 72 × B200 | 8 × H100 |
| Total HBM | 20.7 TB HBM3e | 13.5 TB HBM3e | 640 GB HBM3 |
| FP4 sparse | 2.16 EFLOPS | 1.44 EFLOPS | N/A |
| FP8 throughput | 720 PFLOPS | 720 PFLOPS | 16 PFLOPS |
| NVLink fabric | 130 TB/s | 130 TB/s | 7.2 TB/s (8 GPUs) |
| CPU | 36 × Grace | 36 × Grace | x86 host |
| Networking | ConnectX-8 · 800 Gb/s | ConnectX-7 · 400 Gb/s | ConnectX-7 · 400 Gb/s |
| Spheron availability | Live now | Available | Available |
Per-rack specifications for NVL72 systems. HGX H100 figures are for a standard 8-GPU node for relative scale, since H100 does not come in NVL72 configuration. GB300 specs match the NVIDIA GB300 NVL72 reference design.
Workloads built for GB300
Trillion-Parameter Pre-Training
Each B300 GPU has 288 GB HBM3e. A single 8-GPU node holds 2.3 TB; a full NVL72 rack holds 20.7 TB inside one NVLink domain. 2.16 EFLOPS FP4 sparse per rack cuts wall-clock time on 1T+ parameter training by 4 to 6x versus equivalent H100 capacity. Rent the count that fits your run.
Frontier-Scale FP4 Inference
B300 delivers 1.5x more dense FP4 than B200 (15 vs 10 PFLOPS) and adds 2x attention compute on top. 130 TB/s of NVLink fabric inside an NVL72 rack supports tensor and pipeline parallelism for the largest models without leaving the rack. Smaller rentals fit 70B to 400B parameter inference on 8 to 32 GPUs.
Reasoning and Agentic Workloads
Long reasoning chains generate large KV-cache footprints. Grace CPUs handle orchestration, retrieval, and tool calls in the same coherent address space as the GPUs over 900 GB/s NVLink-C2C, removing PCIe round-trips that slow down agent loops on x86 nodes. Rent from a few GPUs for prototyping up to full racks for production serving.
Fine-Tuning and Post-Training
Rent the GPU count that matches your run. DPO and GRPO on 70B models fit on a single 8-GPU node. Large-scale RL with rollouts, value heads, and reference models benefits from full-rack co-location inside one NVLink domain. Persistent NVMe per chassis handles checkpoint streaming at any scale.
When to pick GB300
Pick GB300 if
You want Blackwell Ultra. We have 1024+ GPUs available to rent at any quantity, from a single 8-GPU node to multi-rack deployments. Models too large for one node sit inside an NVL72 rack with 20.7 TB of unified HBM3e at 130 TB/s. Rent the count you actually need.
Pick B300 (per GPU) instead if
You want per-minute billing with no commitment. B300 SXM6 is available on Spheron on a per-GPU or 8-GPU node basis with on-demand and spot pricing. For most 70B to 200B parameter inference and fine-tuning jobs that don't need a sales conversation, B300 is the simpler call.
Pick GB200 instead if
Budget matters and you don't need 288 GB per GPU. GB200 has 192 GB per GPU versus 288 GB on GB300, with identical NVLink fabric and Grace CPU layout. For sub-1T parameter workloads, GB200 is often the better price-to-performance pick.
Pick R100 if you can wait
R100 Rubin is available H2 2026 with 22 TB/s per-GPU bandwidth (2.75x B300) and 50 PFLOPS FP4 (3.33x B300). For new training runs with flexible timelines, R100 is the higher-ceiling option. If you need to start now, GB300 is live today.
Other GPUs on Spheron
For workloads that don't need GB300 specifically, Spheron also offers B300, B200, and H200 on per-minute billing with no commitments.
Related resources
NVIDIA GB300 NVL72 Architecture Guide
How the NVL72 rack pairs 72 B300 GPUs with 36 Grace CPUs, 20.7 TB of unified HBM3e, NVLink 5 fabric, and what workloads benefit from each scale.
NVIDIA B300 GPU: Specs, Pricing, and Workloads
Single-GPU and 8-GPU B300 SXM6 access on Spheron. 288 GB HBM3e, FP4 Transformer Engine, and per-minute billing with no commitments.
Vera Rubin NVL72: The Next-Gen System
How NVL72 evolves from Blackwell to Rubin. HBM4, NVLink 6, and the upgrade path for teams running GB300 today.