From Exascale to Everyscale™ — AI inferencing at scale
Optimized for AI inferencing at scale with 8x NVIDIA RTX PRO 6000 Blackwell GPUs and AMD EPYC™ processors.Massive memory bandwidth with 1.5TB DDR5 for large inferencing models.Ultra-fast storage with NVMe E3.S and M.2 drives for low-latency data access.PCIe Gen5 architecture for maximum GPU and networking throughput.Advanced cooling with high-performance fans and optional Lenovo Neptune™ hybrid liquid cooling.Enterprise-grade security including TPM 2.0, intrusion detection, and firmware resilience.
Key Features
The SR675i V3 delivers unparalleled GPU density with support for up to 8 NVIDIA RTX PRO 6000 Blackwell GPUs. This configuration accelerates inferencing workloads for computer vision, NLP, and generative AI, enabling real-time insights across industries.
Dual AMD EPYC™ 9535 processors combined with PCIe Gen5 architecture ensure maximum throughput for inferencing models. With 24 DDR5 DIMM slots and high-speed NVMe storage, SR675i V3 is built for performance at scale.
High-performance fans and optional Lenovo Neptune™ hybrid cooling keep GPUs and CPUs operating efficiently under heavy inferencing loads. Rugged design ensures reliability for enterprise and HPC environments.
Specifications
| Form Factor |
|
| Processor |
|
| Memory |
|
| Storage |
|
| GPU |
|
| Expansion Slots |
|
| Networking |
|
| Power Supply |
|
| Cooling |
|
| Security |
|
| Management |
|
| OS Support |
|
| Warranty |
|
| Datasheet |
|
HPE ProLiant DL380 Gen11 Standard Heat Sink Kit (P49145-B21)
HPE MR408i-o Gen11 x8 Lanes 4GB Cache OCP SPDM Storage Controller (P58335-B21) 






There are no reviews yet.