• NVIDIA Ampere architecture CUDA cores: Double-speed processing for single-precision (FP32) operations and improved power efficiency for graphics and computer workflows.
• Second-generation RT cores: Up to 2x throughput of previous generation for speeding workloads such as photorealistic rendering, architectural design, and virtual prototyping.
• Third-generation Tensor Cores: Tensor Float-32 (TF32) precision; provides dramatically increased training throughput to accelerate AI and data science model training.
• PCIe Express Gen 4: Increased bandwidth improves data-transfer speeds from CPU memory.
• Data center security: Secure and measured boot with primary root-of-trust within the GPU helps certify that firmware isn't tampered with or corrupted.

Key Features | Increased Performance to Solve Problems Faster

The NVIDIA Accelerators for HPE servers improve computational performance, dramatically reducing the completion time for parallel tasks, offering quicker time to solutions.

Simplified Integration and Management

NVIDIA Accelerators for HPE are designed to simplify integration with a scalable and highly-configurable modular architecture.

NVIDIA accelerators can be configured and monitored by HPE Insight Cluster Management Utility (CMU). HPE Insight CMU monitors and displays GPU health and temperature, as well as installs and provisions the GPU drivers and CUDA software.