Intel Xeon 6979P Granite Rapids 120-Core Processor

The Intel Xeon 6979P is a flagship single-socket processor built on the Granite Rapids microarchitecture, delivering 120 cores and 120 threads within Intel's high-efficiency P-core design. Fabricated on Intel's Intel 3 process node, it integrates a dramatically expanded last-level cache, high-bandwidth memory interface, and PCIe 5.0 connectivity to address the most demanding enterprise compute requirements. The processor supports up to 8-channel DDR5 memory per socket and introduces the Intel Xeon 6 platform's enhanced AI inference acceleration via AMX (Advanced Matrix Extensions), making it suitable for mixed workloads that span traditional enterprise computing and emerging AI-adjacent tasks.

Granite Rapids marks a substantial leap in per-socket throughput for general-purpose workloads. The 6979P operates with a base clock of 2.0 GHz and supports Turbo Boost frequencies up to 3.7 GHz, balancing sustained multi-threaded throughput with burst performance for latency-sensitive tasks. Its 504 MB of total Intel Smart Cache (L3) is among the largest available in any x86 server processor, directly benefiting in-memory databases, caching tiers, and virtualization platforms that rely on reducing DRAM access latency. The processor has a rated TDP of 500W and is designed for deployment in platforms built around the Intel E1S socket.

In enterprise virtualization environments, the 6979P's 120-core count enables extremely high virtual machine density per node, reducing the physical server footprint and associated infrastructure overhead. The platform supports Intel vRAN Boost, Intel Speed Select Technology, and hardware-enforced security features including Intel TME-MK (Total Memory Encryption – Multi-Key) and Intel TDX (Trust Domain Extensions), giving organizations granular control over workload isolation, performance tuning, and data confidentiality across multi-tenant deployments.

Ideal for

• High-density enterprise virtualization and private cloud consolidation, maximizing VM-per-node ratios on VMware vSphere, Red Hat OpenShift, or Microsoft Hyper-V platforms
• Large-scale containerized microservices environments requiring high core counts per node for Kubernetes cluster worker nodes
• In-memory and real-time analytics databases such as SAP HANA, Redis, or VoltDB that benefit from the 504 MB Smart Cache reducing DRAM access latency
• AI inference serving for enterprise models using Intel AMX acceleration to handle concurrent inference requests at scale without dedicated GPU infrastructure
• Software-defined networking and network function virtualization (NFV) deployments that require high throughput CPU-based packet processing
• High-performance computing (HPC) simulation and modeling workloads in financial services, life sciences, and engineering that demand maximum floating-point throughput per rack unit

Technical specifications

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