Gigabyte G292-Z43 4x GPU Server with AMD MI300X OAM (Bare)
Brand: Gigabyte | Category: GPUs
SKU: G292-Z43-AAX1 | Part #: G292-Z43-AAX1 | MPN: G292-Z43-AAX1
Contact for Pricing — Request a Quote
About the Gigabyte G292-Z43 4x GPU Server with AMD MI300X OAM (Bare)
For large-scale AI training and high-performance compute workloads, the Gigabyte G292-Z43 4x GPU Server delivers exceptional density and bandwidth through four AMD Instinct MI300X OAM accelerators. This 2U rackmount system (part number G292-Z43-AAX1) is purpose-built for organizations deploying GPU-accelerated infrastructure at enterprise scale. The four AMD Instinct MI300X OAM accelerators provide 768 GB of total HBM3 GPU memory—192 GB per accelerator—with each GPU delivering 5.3 TB/s of memory bandwidth. GPU-to-GPU connectivity via AMD Infinity Fabric ensures efficient multi-GPU communication, while the system supports dual-socket AMD EPYC 9004 Series (Genoa) processors on the SP5 platform for balanced compute performance alongside GPU acceleration.
The Gigabyte G292-Z43-AAX1 is the choice of AI infrastructure teams, HPC centers, and data center architects who require maximum GPU throughput without compromise. The server includes 24x DDR5 DIMM slots for system memory, eight hot-swap 2.5-inch bays supporting NVMe, SATA, and SAS drives, PCIe 5.0 expansion, and an OCP 3.0 mezzanine slot for high-bandwidth networking. Redundant 3000W Titanium-level power supplies ensure reliability in demanding environments, while the Gigabyte Management Controller with IPMI 2.0 and ASPEED AST2600 BMC provides comprehensive out-of-band management. The system operates in ambient temperatures from 10°C to 35°C and is compatible with AMD ROCm software, positioning it for flexible deployment across diverse AI and compute clusters.
The Gigabyte G292-Z43 arrives as a bare system—operating system not included—allowing you to standardize your software stack and deployment methodology. For qualified enterprises ready to evaluate this GPU server for pilot deployments or production procurement, contact Omnixon Global to submit your detailed request for quotation and technical specifications discussion.
Technical Specifications
| Brand | Gigabyte |
| Category | GPUs |
| SKU | G292-Z43-AAX1 |
| Part Number | G292-Z43-AAX1 |
| Condition | New |
| Manufacturer Part Number | G292-Z43-AAX1 |
| Form Factor | 2U Rackmount |
| GPU Accelerators | 4x AMD Instinct MI300X OAM |
| GPU Memory per Accelerator | 192 GB HBM3 |
| Total GPU Memory | 768 GB HBM3 (4x 192 GB) |
| GPU Memory Bandwidth per Accelerator | 5.3 TB/s |
| GPU Interconnect | AMD Infinity Fabric (GPU-to-GPU) |
| CPU Support | AMD EPYC 9004 Series (Genoa), Dual Socket SP5 |
| System Memory Type | DDR5 |
| System Memory Slots | 24x DIMM slots |
| Storage Bays | 8x 2.5-inch Hot-Swap NVMe/SATA/SAS |
| PCIe Expansion | PCIe 5.0 |
| Network Expansion | OCP 3.0 Mezzanine slot; PCIe expansion slots for high-bandwidth NICs |
| Power Supply | Redundant 3000W Titanium-level PSUs |
| Management | Gigabyte Management Controller (IPMI 2.0, ASPEED AST2600 BMC) |
| Operating System | None (Bare / OS-not-included) |
| GPU Software Stack | AMD ROCm (compatible) |
| Chassis Width | Standard 19-inch rack compatible |
| Operating Temperature | 10°C to 35°C |
Frequently Asked Questions about Gigabyte G292-Z43 4x GPU Server with AMD MI300X OAM (Bare)
What server platforms accept the Gigabyte G292-Z43 4x GPU Server with AMD MI300X OAM (Bare)?
Reference servers include Dell PowerEdge XE9680 / XE9712, HPE Cray XD670, Lenovo ThinkSystem SR685a / SR675 V3, Supermicro AS-A21GE / SYS-821GE, Gigabyte G593 / G894, ASUS ESC. Share your target platform in the RFQ and we will confirm chassis-to-GPU compatibility and recommended NIC pairing.
How long is the lead time on AI GPUs?
Highly model-dependent. L40S / RTX-class: typically 3-6 weeks. H100/H200/B200 in SXM form factor: 12-16 weeks for whole-platform allocations. We quote genuine-channel ETAs only — no grey-market promises.
Do you supply matched networking (Quantum InfiniBand / Spectrum-X)?
Yes — Omnixon stocks the full NVIDIA networking lineup (Quantum-2 / Quantum-X InfiniBand, Spectrum-X Ethernet, ConnectX NICs, BlueField DPUs) so we can quote a complete training-cluster BOM, not just the GPUs.
Can you help with NVIDIA AI Enterprise licensing?
Yes. We hold genuine-channels for NVIDIA AI Enterprise software subscriptions. Add it to your RFQ and we quote node-aligned licensing along with the hardware.