| Sign In | Join Free | My infospaceinc.com |
|
Model Number : R5300 G5
Price : Negotiable
Memory Capacity : 64GB
Brand Name : H3C
Packaging Details : Carton Packaging
Dimensions : 90X60X30 cm
Payment Terms : T/T
Pbox Switching : 598Gbps
Power Supply : 1+1 Redundancy power supply
Place of Origin : CHINA
Switching Capacity : 176Gbps
Supply Ability : 200pcs
MOQ : 1pc
Delivery Time : 5-7 work days
Productname : H3C Server
Box Switching : 598Gbps
Warranty : 3 Years
Flexible GPU Modules | 32x DDR4 DIMMs (12TB Max) | PCIe 4.0 | Up to 200GE Networking
The H3C UniServer R5300 G5 is a groundbreaking 4U dual-socket heterogeneous computing server for deep learning training, inference, HPC, and big data analytics. As the successor to the R5300 G3, it delivers significantly enhanced CPU/GPU specifications and modular architecture.
| Category | Specification |
|---|---|
| Form Factor | 4U rackmount |
| Processors | 2x Intel Xeon 3rd Gen Scalable (40 cores/CPU max, 270W TDP) |
| Chipset | Intel C741 |
| Memory | 32x DDR4 DIMMs, up to 3200MT/s, 12TB max; 16x Optane PMem 200 series |
| Storage | 24x or 12x 2.5/3.5" SAS/SATA/NVMe drives |
| Cooling | 8x 8056 fans, N+1 hot-plug redundant |
| Power Supply | 4x 2000W/2400W Platinum, N+N redundancy |
| Management | HDM agent-free, H3C FIST |
| Operating Temp | 5°C to 35°C |
| Dimensions | 174.8 x 447 x 807 mm |
| Warranty | 3 years, 5×9, next business day |
| Module | Configuration | Best For |
|---|---|---|
| 8-GPU Module | 8 double-width GPUs + 4 single-width slots | Deep learning training, HPC |
| 20-GPU Module | 20 single-width GPUs | High-throughput inference |
| HGX NVLINK | 4x A800 GPUs (600GB/s NVLINK) | AI training |
Floating-point performance: Up to 52% improvement over previous generation
Core count: Up to 42% increase
Memory bandwidth: 60% improvement (3200MHz vs previous)
| Feature | Specification |
|---|---|
| PCIe Slots (8-GPU) | 8x double-width x16 + 4x single-width x16 + OCP 3.0 + RAID |
| PCIe Slots (20-GPU) | 20x single-width x16 + OCP 3.0 + RAID |
| Network | OCP 3.0 + optional 1/10/25/40/50/100/200GE |
| Use Case | Benefit |
|---|---|
| Deep Learning Training | Up to 8 double-width GPUs, 12TB memory |
| Deep Learning Inference | Up to 20 single-width GPUs for high throughput |
| HPC | PCIe 4.0, 3200MHz memory |
| Big Data Analytics | Massive storage + 200GE networking |
| Benefit | Description |
|---|---|
| Superior Performance | 52% FP improvement, 60% memory bandwidth |
| Unmatched Flexibility | 3 GPU modules, 1:4/1:8 CPU-to-GPU ratios |
| Future-Ready | PCIe 4.0, up to 200GE |
| Enterprise Reliability | N+N power, N+1 fans, redundant design |
| Proven Track Record | Industry-recognized AI computing platform |
The H3C UniServer R5300 G5 combines flexible GPU modules (8/20 GPUs or HGX NVLINK), 32 DDR4 DIMMs (12TB max, 3200MT/s), PCIe 4.0, and up to 200GE networking in a 4U form factor. With modular architecture, enterprise-grade redundancy, and comprehensive management, it delivers the performance, scalability, and reliability required for AI training, inference, HPC, and big data analytics workloads.
|
|
AI Training HPC H3C Server UniServer R5300 G5 4U Rack Heterogeneous Computing Platform Images |