@Fluence Updates Its Compute Network With Two New NVIDIA B300 Clusters
Two NVIDIA HGX B300 clusters are now open for reservation through Fluence, bringing more high-performance GPU capacity to its compute network.
Here’s the interesting part:
🇺🇸 Texas: 10 liquid-cooled nodes, with 8 B300 GPUs per node. Each GPU has 288 GB of HBM3E memory, giving each node 2.3 TB of GPU memory. The cluster can be ready within one week of signing.
🇺🇸 Nevada: 6 air-cooled nodes using Aivres KR9288X3 systems with HGX B300 GPUs, expected to be ready on January 1, 2027.
Both options offer dedicated bare-metal infrastructure, full root access, 2 TB of system memory, and 30.72 TB of NVMe storage.
What I find interesting here is the practical side of the update. These aren't just new hardware specifications to read about. Teams can reserve the infrastructure for their own workloads and have direct control over their systems.
As AI workloads become more demanding, I think the infrastructure powering them deserves just as much attention as the applications everyone sees.
Fluence is continuing to build in that space, and I'm curious to see how its compute network develops from here.
#Fluence #AI #DePIN
Two NVIDIA HGX B300 clusters are now open for reservation through Fluence, bringing more high-performance GPU capacity to its compute network.
Here’s the interesting part:
🇺🇸 Texas: 10 liquid-cooled nodes, with 8 B300 GPUs per node. Each GPU has 288 GB of HBM3E memory, giving each node 2.3 TB of GPU memory. The cluster can be ready within one week of signing.
🇺🇸 Nevada: 6 air-cooled nodes using Aivres KR9288X3 systems with HGX B300 GPUs, expected to be ready on January 1, 2027.
Both options offer dedicated bare-metal infrastructure, full root access, 2 TB of system memory, and 30.72 TB of NVMe storage.
What I find interesting here is the practical side of the update. These aren't just new hardware specifications to read about. Teams can reserve the infrastructure for their own workloads and have direct control over their systems.
As AI workloads become more demanding, I think the infrastructure powering them deserves just as much attention as the applications everyone sees.
Fluence is continuing to build in that space, and I'm curious to see how its compute network develops from here.
#Fluence #AI #DePIN