G893-ZX1-AAX2 Gigabyte Dual EPYC 9004 8U 6NG893ZX1DR000ACX2 Server Barebone
Do you have a major deal consisting of various components? Then you're at the right place!
We offer you the best prices for your purchase:
- - If you already have an account with us, please log in. If you don't have an account, you can apply for one here.
- - simply collect your required items in the desired quantity in CART, and click the 'GET QUOTATION' button. and you're done!
- - You will receive your offer within the next 24 hours on working days.
Gigabyte G Series
Do you have questions about this product 6NG893ZX1DR000ACX2?
G893-ZX1-AAX2 Gigabyte
High-Density 8U GPU-Accelerated Rack Server for MI325X OAM AI/HPC featuring the G893-ZX1-AAX2 Gigabyte architecture
The G893-ZX1-AAX2 Gigabyte platform is an 8U dual-socket AMD EPYC SP5 GPU server engineered to consolidate large-scale accelerator capacity into a single chassis for AI training, HPC compute, and high-throughput inference. With 8x AMD Instinct MI325X OAM accelerators, the design prioritises direct GPU-to-CPU and GPU-to-I/O efficiency while keeping serviceability aligned with enterprise rack operations.
Compute is anchored by 2x AMD EPYC 9004/9005 (Genoa/Turin) processors, enabling a high-core-count host layer for data preparation, orchestration, and CPU-bound stages in heterogeneous pipelines. The memory subsystem uses DDR5 ECC RDIMM across 24 DIMM slots with 12 channels per CPU, supporting up to 6TB for large model states, simulation datasets, and in-memory batching. PCIe Gen5 I/O is central to the G893-ZX1-AAX2 Gigabyte layout, providing 12x PCIe Gen5 x16 FHHL slots for high-bandwidth fabrics and storage controllers, while front-access 8x 2.5" hot-swap NVMe bays deliver low-latency Gen5 local data staging. Internal M.2 slots support OS boot separation for cleaner lifecycle management.
Why Choose the G893-ZX1-AAX2 Gigabyte?
The G893-ZX1-AAX2 Gigabyte system delivers a balanced density-to-performance profile by pairing 8x MI325X OAM GPUs with dual EPYC SP5 CPUs, allowing enterprises to scale accelerator throughput without fragmenting clusters across multiple smaller nodes. Redundant 80 PLUS Titanium power, hot-swap service points, and BMC-led remote operations support predictable uptime and standardised deployment in AI and HPC racks.
Gigabyte focuses on platform engineering that integrates modern management (ASPEED AST2600 with IPMI/HTML5 KVM), deterministic PCIe Gen5 expansion, and datacentre-ready thermals and power design to meet continuous-duty requirements.
Key Applications:
- Large-scale AI model training
- GPU-accelerated HPC simulation
- Batch AI inference
- Data preparation and feature engineering
- Accelerated research computing
At Happyware - Server Hardware you benefit from Non-Binding Offers tailored to the G893-ZX1-AAX2 Gigabyte configuration and deployment goals, helping you align GPU density, networking, and storage choices to your cluster design. For long-running AI/HPC environments, Extended Warranty options up to 6 years add planning security across refresh cycles.
Technical details of G893-ZX1-AAX2 Gigabyte
How long is the Warranty and how you can get technical support for G893-ZX1-AAX2 Gigabyte
| Warranty: | 36 months (part) |
|---|---|
| RMA Url: | https://www.gigabyte.com/Support/Consumer/Warranty |
| Description: | Items must be returned first. The manufacturer does not send replacement items in advance. Delivery times are not guaranteed, and on-site service is not offered as standard. |
| Support Portal: | https://support.happyware.com |
|---|
Customer reviews about G893-ZX1-AAX2 Gigabyte
Login
- Express shopping
- Save your data and settings
- Order overview and shipping information
- Manage your newsletter subscription
Deals for G893-ZX1-AAX2 Gigabyte
We don’t have any deals for this product yet.