NVIDIA HGX H200
Eight Hopper GPUs with the larger memory. Air cooled, one node per rack in an existing hall.
- 8 GPUs per node
- 1.1 TB GPU memory
- Air
- Available Q1 2027
Configurations
Pick the footprint that fits. Every one is sold on the same contract.
- 8 GPUsOne node
- The smallest unit. One node per rack in an existing Santiago hall.
- 16 to 248 GPUsTwo to thirty-one nodes
- One fabric, one cluster.
- 256 GPUs and moreThirty-two nodes and up
- A row of racks, no special cooling.
Specifications
- GPUs
- 8 Hopper per node
- GPU memory
- 1.1 TB, 141 GB HBM3e per GPU
- NVLink
- 900 GB/s per GPU, 7.2 TB/s in the node
- CPUs
- 2 x86 CPUs
- System memory
- 1.5 TB or more
- Networking
- BlueField-3 SuperNICs, up to 400 Gb/s per GPU
- Power
- About 10.2 kW per node; up to 700 W per GPU
- Cooling
- Air
- Fabric at Fulmis
- InfiniBand or Spectrum-X Ethernet, sized to the workload
- Software
- Slurm and managed Kubernetes
- Location
- A data centre in the Santiago Metropolitan Region
- Available
- Q1 2027
Performance
Per node, as NVIDIA publishes it.
- FP8 Tensor, per GPU
- 3,958 TFLOPS
- FP16 / BF16 Tensor, per GPU
- 1,979 TFLOPS
- TF32 Tensor, per GPU
- 989 TFLOPS
- GPU memory
- 1.1 TB HBM3e, 4.8 TB/s per GPU
- NVLink bandwidth
- 7.2 TB/s
- Max GPU power
- Up to 700 W per GPU, configurable
- Node power
- About 10.2 kW
NVIDIA's published figures. Tensor Core numbers carry sparsity unless the row says dense; H100, H200 and RTX PRO figures are per GPU because NVIDIA publishes no node total. Power figures are the ones we reserve and protect to, with NVIDIA's own where it publishes one.
How we deploy it
HGX H200 is the eight-GPU Hopper node with the larger HBM. It is air cooled at about 10 kW, which places one node per rack in an existing Santiago hall with no special cooling, and it is the platform for a team whose data has to stay in Chile and whose workload fits in eight GPUs.
A node is the smallest unit. It is sold on the same three to five year contract as every other platform.
Fulmis places the purchase order, or takes delivery of the fleet you already own. Either way we bring it into Chile, commission it against the vendor baseline and operate it for the term. You reserve the capacity, or your own hardware; the workload and the data stay yours.
Terms
- Term
- Three to five years
- Price
- Fixed for the term, in the term sheet
- Commitment
- Take-or-pay
- Prepayment
- None to 25% at signature
- Smallest block
- One node, 8 GPUs
- Start
- Q1 2027, first slot
Other platforms
| Platform | GPUs | GPU memory | NVLink | Power | Cooling | Smallest block | Available |
|---|---|---|---|---|---|---|---|
| GB300 NVL72Rack scale | 72per rack | 20 TB288 GB per GPU | 130 TB/s1.8 TB/s per GPU | 132 to 140 kW155 kW peak | Liquid | 1 rack, 72 GPUs | Q1 2027 |
| GB200 NVL72Rack scale | 72per rack | 13.4 TB186 GB per GPU | 130 TB/s1.8 TB/s per GPU | 120 to 132 kWper rack | Liquid | 1 rack, 72 GPUs | Q1 2027 |
| HGX B3008-GPU node | 8per node | 2.3 TB288 GB per GPU | 14.4 TB/s1.8 TB/s per GPU | About 14 kWper node | Air or liquid | 1 node, 8 GPUs | December 2026 |
| HGX B2008-GPU node | 8per node | 1.4 TB180 GB per GPU | 14.4 TB/s1.8 TB/s per GPU | About 14.3 kWper node | Air or liquid | 1 node, 8 GPUs | December 2026 |
| HGX H1008-GPU node | 8per node | 640 GB80 GB per GPU | 7.2 TB/s900 GB/s per GPU | About 10.2 kWper node | Air | 1 node, 8 GPUs | Q1 2027 |
| RTX PRO 6000 Blackwell8-GPU node | 8per node | 768 GB96 GB per GPU | PCIe Gen 5no NVLink | 5.4 to 6.7 kWper node | Air | 1 node, 8 GPUs | Q1 2027 |
| Vera RubinRack scale | 72per rack | HBM4not yet published | Not published | Not published | Liquid | 1 rack | On request |