GIGABYTE R283-Z90 2U Rackmount Server (Dual AMD EPYC 9005)

The GIGABYTE R283-Z90 is a 2U rackmount server engineered around the dual-socket AMD EPYC 9005 series (5th Generation "Turin") platform, delivering a substantial leap in x86 core density and memory bandwidth for enterprise data center environments. Built on the AMD EPYC 9005 architecture with Zen 5 cores, the platform supports up to two processors with a combined maximum TDP envelope suited to high-throughput compute tasks, backed by 12-channel DDR5 memory per socket for exceptional memory bandwidth across 24 DIMM slots supporting up to 3TB of registered ECC DDR5 capacity.

The R283-Z90 follows GIGABYTE's established R283 chassis lineage, optimised for the SP5 socket transition to Turin CPUs, and retains a flexible storage and I/O configuration. The system accommodates multiple NVMe and SATA/SAS drive bays in a front-accessible 2U footprint, paired with a full-length PCIe 5.0 expansion slot complement to support high-bandwidth accelerators, networking adapters, and storage controllers. Dual redundant 80 PLUS Titanium-rated power supplies maintain high efficiency under variable load conditions, while the integrated GIGABYTE Server Management (GSM) BMC with IPMI 2.0 and Redfish API support enables out-of-band management and integration with modern infrastructure automation toolchains.

Targeted at enterprise IT teams deploying general-purpose compute, virtualisation hosts, in-memory databases, and scale-out HPC workloads, the R283-Z90 represents a balanced 2U platform that maximises core count, memory capacity, and I/O throughput within standard rack power and space envelopes. Its refreshed EPYC 9005 readiness makes it a relevant long-cycle investment for organisations standardising on AMD's current-generation server ecosystem.

Ideal for

• Large-scale virtualisation hosts running VMware vSphere, Microsoft Hyper-V, or KVM hypervisors requiring high vCPU-to-core ratios and large memory footprints
• Enterprise database servers for Oracle, Microsoft SQL Server, or PostgreSQL deployments benefiting from high DDR5 memory bandwidth and capacity
• High-performance computing (HPC) cluster nodes for scientific modelling, financial risk simulation, and engineering analysis workloads
• Cloud-native infrastructure nodes supporting containerised workloads orchestrated via Kubernetes, leveraging high core counts and PCIe 5.0 for fast networking
• AI inference serving platforms where CPU-side pre/post-processing demands and high memory bandwidth complement attached accelerator cards
• Software-defined storage controller nodes using NVMe over Fabrics or Ceph, taking advantage of PCIe 5.0 bandwidth and abundant drive bay capacity

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.