Gigabyte Enterprise · GPU Servers
Gigabyte G4L3-ZD1-LAX5
HPC/AI Server - AMD EPYC™ 9005/9004 - 4U DP NVIDIA HGX™ B200 DLC
The Gigabyte G4L3-ZD1-LAX5 is one of the G4L3 series 4U GPU servers. It comes with 2 × AMD EPYC processors, 24 DIMM slots, 8 x 2.5" drive bays, and direct liquid cooling (DLC). It supports 8 × NVIDIA HGX B200 GPUs. As the authorized distributor of Gigabyte Enterprise in Türkiye, TopTel delivers this model with project-specific configuration, local stock and QPS deployment support.
Highlights
- Liquid-cooled NVIDIA HGX™ B200
- CPU+GPU direct liquid cooling solution with leak detection
- 900GB/s GPU-to-GPU bandwidth with NVIDIA NVLink™ and NVSwitch™
- Dual AMD EPYC™ 9005/9004 Series Processors
- 12-Channel DDR5 RDIMM, 24 x DIMMs
- Dual ROM Architecture
- Compatible with NVIDIA® BlueField®-3 DPUs and NVIDIA ConnectX®-7 NICs
- 2 x 10Gb/s LAN ports via Intel® X710-AT2
- 2 x M.2 slots with PCIe Gen3 x4 and x1 interface
- 8 x 2.5" Gen5 NVMe hot-swap bays
- 4 x FHHL dual-slot PCIe Gen5 x16 slots
- 8 x FHHL single-slot PCIe Gen5 x16 slots
- 4+4 3000W 80 PLUS Titanium redundant power supplies
Technical Specifications
| Form Factor | 4U |
| CPU Series | AMD EPYC |
| CPU Type | AMD EPYC 9004, AMD EPYC 9005 |
| CPUs | 2 |
| CPU Socket | Socket SP5 |
| Chipset | AMD® SoC |
| DIMM Slots | 24 |
| Drive Bays | 8 x 2.5" |
| GPU Type | NVIDIA HGX B200 |
| GPUs | 8 |
| LAN | 2 × 10 Gb/s |
| Cooling Type | Direct Liquid Cooling |
| PSU | 8 x 3000 W |
| Platform | x86 Server |
| Application | AI Platform, High-Performance Computing Server, AI Training Server, AI Inference Server |
| Processor Vendor | AMD |
| Series | G Series |
| Max TDP (cTDP) | 500 W |
| GPU Series | NVIDIA HGX |
Use cases
- Large language model (LLM) training and fine-tuning
- AI inference servers (RAG, chatbots, vision)
- HPC and scientific computing (CFD, molecular dynamics)
- Render farms and media processing
- vGPU in VDI environments
Sourcing the G4L3-ZD1-LAX5 from TopTel
- Local stock and fast delivery (Istanbul)
- Project-specific configuration: memory, disk, GPU, NIC selection
- Deployment with QPS engineering, BIOS/BMC configuration and commissioning
- On-site service and RMA management in Turkey
- Validated with Sangfor HCI as part of the GigaX integrated system
Other models in the G4L3 series
Frequently asked questions
Which GPUs can this model use?
The GPU types listed in the technical table are supported; other GPUs are confirmed against the Gigabyte QVL.
How many kW of power are required?
Fully loaded power draw depends on the PSU capacity; TopTel supports rack planning.
What is the lead time?
In-stock models ship in 1–2 weeks; custom configurations in 4–8 weeks.
Get a quote for the G4L3-ZD1-LAX5
We respond within 1 business day with configuration, stock and lead time.