HPE ProLiant DL380a Gen11 vs Lenovo ThinkSystem SR670 V2
Side-by-side specifications from the vendors' own documents. The HPE ProLiant DL380a Gen11 is a 2U dual-socket accelerator-optimized rack server designed specifically for enterprise artificial intelligence, machine learning inference, and deep learning training. The ThinkSystem SR670 V2 is Lenovo's modular 3U two-socket high-density GPU server engineered for artificial intelligence, machine learning deep training, high-performance computing (HPC), and graphical rendering clusters.
| HPE ProLiant DL380a Gen11 | Lenovo ThinkSystem SR670 V2 | |
|---|---|---|
| Brand | HPE | Lenovo |
| Status | current | end-of-sale |
| Released | Jun 2023 | Apr 2021 |
| Form factor | 2U rack | 3U rack |
| Cpu sockets | 2 | 2 |
| Cpu platform | 4th and 5th Gen Intel Xeon Scalable processors up to 64 cores per socket | Dual 3rd Gen Intel Xeon Scalable processors (Ice Lake-SP) up to 270W TDP |
| Memory slots | 32 | 32 |
| Memory type | DDR5-4800 / DDR5-5600 SmartMemory RDIMM | DDR4-3200 RDIMM / 3DS RDIMM |
| Memory max (GB) | 8,192 | 4,096 |
| Drive bays | Up to 8 SFF NVMe PCIe SSDs or EDSFF E3.S | Up to 8x 2.5-inch SFF hot-swap SAS/SATA/NVMe drive bays |
| Pcie slots | Up to 4 double-width or 8 single-width PCIe 5.0 GPU slots | Up to 8 PCIe 4.0 x16 slots for GPUs plus 4 PCIe 4.0 x16 slots for high-speed NICs/InfiniBand |
| Gpu support | Optimized for up to 4 NVIDIA L40S, L4, or H100 NVL GPUs | Up to 8 double-wide PCIe GPUs (such as NVIDIA A100/A40) or 4-way NVIDIA HGX A100 NVLink baseboard with Lenovo Neptune hybrid liquid cooling |
| Power supply | Up to 4x 1600W or 1800W Flex Slot Titanium hot-plug | Up to 4 redundant 1800W / 2400W AC 80 PLUS Platinum hot-swap power supplies |
| Network onboard | — | Dual OCP 3.0 slots plus dedicated 1GbE XClarity management port |
Buying or selling a DL380a Gen11 or SR670 V2?
New, refurbished or used. Tell us the quantity and condition and we connect you with a vetted reseller in your region, usually within one working day.