AMD EPYC™ 9006 Series Server CPUs
The AMD EPYC™ 9006 Series is designed for the next generation of AI-first data centers. Based on the latest AMD “Zen 6” and “Zen 6c” architectures, the new server processor family combines high core density, fast CPU-to-GPU communication and high memory bandwidth for demanding AI, cloud, enterprise and HPC workloads.
With configurations offering up to 256 cores and 512 threads, EPYC 9006 processors provide the compute density required to run more applications and AI agents efficiently across modern server infrastructures.
Designed for the Agentic AI Era
Modern AI infrastructure requires more than powerful GPU accelerators. CPUs remain essential for orchestration, data processing, retrieval, tool execution and keeping accelerators supplied with data.
AMD EPYC 9006 Series processors are designed to support three important roles within AI infrastructure:
- Agent Sandbox CPU – run large numbers of AI agents efficiently
- AI Host Node CPU – support GPU accelerators with high-frequency compute, memory bandwidth and fast I/O
- General-Purpose CPU – handle enterprise, orchestration and mixed workloads
The combination of high core and thread density allows data centers to scale AI workloads while making efficient use of available rack space and power.
Built on AMD “Zen 6” and “Zen 6c”
The EPYC 9006 Series combines “Zen 6” and “Zen 6c” CPU cores to address different workload requirements.
High-frequency cores are suited to latency-sensitive applications and AI host nodes, while high-density core configurations enable large numbers of workloads and AI agents to run in parallel.
Key platform capabilities include:
- Up to 256 CPU cores
- Up to 512 threads
- PCI Express® Gen 6 connectivity
- High memory bandwidth
- Up to 16 memory channels on supported platforms
- Select processors with boost frequencies up to 5 GHz
A Broad Server CPU Portfolio
The AMD EPYC 9006 family provides several platform options designed for different data center requirements.
EPYC 9006 SP7
Designed for high core density and performance across AI, cloud, virtualization and general-purpose workloads.
EPYC 9006 SP8
Provides a balanced platform for enterprise applications, virtualization and mixed business workloads.
EPYC 9006X SP7
Designed for HPC, simulation, modelling and memory-intensive technical computing workloads with enhanced cache and bandwidth.
EPYC 9006 LP
Optimized specifically for AI host-node deployments and efficient utilization of GPU accelerators.
Built for Modern Data Center Workloads
The AMD EPYC 9006 Series is designed to support a broad range of applications, including:
- Agentic and generative AI infrastructure
- GPU-accelerated AI servers
- Cloud computing
- Virtualization
- Enterprise applications
- High-performance computing
- Simulation and modelling
- Data-intensive workloads
This broad portfolio allows organizations to select the processor configuration that best matches the performance, efficiency and density requirements of each server workload.
A CPU Foundation for AI-First Infrastructure
As AI workloads become more complex, the CPU plays an increasingly important role in keeping data moving efficiently between applications, memory and GPU accelerators.
The AMD EPYC 9006 Series combines high core density, modern memory capabilities and PCIe Gen 6 connectivity to provide a scalable CPU foundation for AI-first data centers, while continuing to support traditional enterprise, cloud and HPC workloads.
AMD EPYC 9006 Servers at DALCO
Looking to build or upgrade your server infrastructure with AMD EPYC™ 9006 Series processors?
Contact DALCO to find the right server configuration for your AI, HPC, virtualization and enterprise workloads.