| Availability: | |
|---|---|
| Quantity: | |
Xeon 6767p
Xeon 6767P is the high‑core‑count member of Intel Xeon 6 Granite Rapids series, manufactured on Intel3 process node. Equipped with 64 physical cores and 128 threads via Intel Hyper‑Threading technology. Base frequency 2.4GHz, maximum turbo frequency up to 3.9GHz, all‑core boost 3.6GHz, huge 336MB shared cache, default TDP 350W, supports Intel SST‑PP multi‑mode performance profile switching for server / compute mode adjustmentIntel.
It supports 8‑channel DDR5‑6400 and high‑speed MRDIMM‑8000 ECC registered memory, maximum memory capacity up to 4TB. Provides 88 PCIe 5.0 lanes for NVMe SSD, network card and accelerator expansion. Four 24GT/s UPI links ensure stable dual‑socket 2‑processor interconnection. Built‑in Intel AMX matrix acceleration, QAT quick assist, TDX trusted domain extensions, improves AI inference, data compression and hardware‑level data securityIntel.
This CPU adopts FCLGA4710 socket. Widely deployed in high‑performance computing HPC, large‑scale virtualization cluster, enterprise database server and high‑end data center scenarios. High‑power dedicated server cooling solution is required for stable full‑load operation
1. 64‑Core 128‑Thread High‑Density Dual‑Socket Enterprise Server Processor
The Intel Xeon 6767P belongs to the Xeon 6 Granite Rapids‑SP family built on Intel 3 process. Equipped with 64 physical cores and 128 SMT multi‑threads, this high‑end data‑center CPU delivers massive multi‑thread throughput for heavy‑duty workloads. It supports single‑socket and dual‑socket FCLGA4710 platforms, perfectly suited for large‑scale virtualization, HPC simulation, AI inference, high‑concurrency cloud and mission‑critical enterprise computing scenarios。
2. 2.4GHz Base / 3.6GHz All‑Core Boost / Up to 3.9GHz Max Turbo Frequency & Massive 336MB L3 Cache
Running at 2.4GHz base frequency, 3.6GHz all‑core boost and up to 3.9GHz maximum turbo frequency. It supports Intel SST‑PP speed‑select performance profile, switching between Server Mode and Compute Mode for flexible base‑frequency tuning under 350W TDP envelope. Built‑in 336MB shared L3 cache drastically reduces memory access latency, accelerates virtual machine scheduling, large‑batch data processing and AI inference tasks, maintaining stable throughput under long‑term heavy data‑center loadsintel.eu。
3. 8‑Channel DDR5 / MRDIMM Memory & Comprehensive PCIe 5.0 Expansion
Support 8‑channel DDR5‑6400 ECC memory and MRDIMM‑8800, maximum memory capacity up to 4TB per socket, delivering ultra‑high memory bandwidth for memory‑hungry enterprise workloads. Offering 88 PCIe 5.0 lanes, compatible with high‑speed NVMe SSD, 200G network adapters and multi‑GPU / AI accelerator cards. Adopt FCLGA4710 socket, equipped with 4 UPI links running at 24GT/s for stable dual‑socket interconnection, supporting 24/7 non‑stop data‑center‑grade operationIntel。
4. 350W Nominal TDP, SST‑PP Performance Profile & Full Suite Of Enterprise‑Grade Technologies
Nominal 350W TDP, requires high‑performance server cooling solution for full‑load stable operation. Integrated complete enterprise‑level features: Intel VT‑x / VT‑d hardware virtualization, Intel AMX AI acceleration, AVX‑512 advanced instruction set, QAT quick assist, DLB dynamic load balancer, ECC error correction and platform security enhancement. It supports massive virtual machine deployment, container orchestration and large‑scale AI inference workloads. Note: This processor has no integrated graphics outputintel.eu。
5. Datacenter‑Grade IC Reliability & Broad Enterprise Compatibility
Manufactured under strict Intel server‑grade specifications, Tcase maximum operating temperature 81°C, DTS max temperature 101°C, designed for long‑term round‑the‑clock continuous operation. Widely applied for new high‑end server assembly, hyperscale data‑center construction, hardware refresh and premium refurbished server reconstruction. It is a flagship high‑core‑count Xeon 6 compute component for modern heavy‑duty enterprise infrastructureIntel。
Item | Specifications |
|---|---|
Part Number | Intel Xeon 6767P |
Product Series | Xeon 6 (Granite Rapids‑SP) Server Microprocessor IC |
Architecture | Granite Rapids‑SP |
Process | Intel 3 |
Core / Thread | 64 Cores / 128 Threads (SMT Enabled) |
Base Frequency | 2.4 GHz |
All‑Core Boost Frequency | 3.6 GHz |
Max Turbo Frequency | Up to 3.9 GHz |
L3 Cache | 336 MB Smart Cache |
Default TDP | 350 W |
SST‑PP Modes | Server Mode / Compute Mode |
Socket | FCLGA4710 (Support 1P / 2P), 4 UPI links @24GT/s |
Memory Support | 8‑Channel DDR5‑6400 / MRDIMM‑8800 ECC, Max 4TB per socket |
PCIe Version | PCIe 5.0, 88 Lanes |
Integrated Graphics | None |
Key Instruction | AMX, AVX‑512, SSE4.2, AES‑NI, QAT, DLB |
Virtualization | Intel VT‑x, VT‑d |
Max Case Temperature | 81°C |
DTS Max Temperature | 101°C |
Core Advantages | Flagship 64C128T massive multi‑thread performance, huge 336MB L3 cache, Intel AMX AI acceleration, SST‑PP flexible performance profile, MRDIMM high‑speed memory support, 4‑link UPI dual‑socket capability, ideal for hyperscale data‑center and AI‑oriented workloads |
Application Grade | Hyperscale Datacenter Servers, High‑Density Virtualization Hosts, HPC Compute Nodes, AI Inference Servers, Large‑Scale Cloud Platforms, Enterprise Hardware Refresh & High‑End Refurbishment |
1. Hyperscale Cloud & High‑Density Virtualization Host Servers
Xeon 6767P serves as flagship compute core for large‑scale cloud servers. 64‑core 128‑thread resources support massive concurrent virtual machines and container instances, maximizing hardware utilization for hyperscale private cloud and multi‑tenant hosting infrastructure.
2. HPC Simulation & Large‑Batch Computing Nodes
Massive multi‑thread capability plus Intel AMX acceleration handle scientific simulation, numerical calculation, media transcoding and large‑scale offline batch‑task scheduling, suitable for high‑end HPC cluster construction.
3. AI Inference & Accelerator‑Attached Server Platforms
Optimized for AI inference workloads, capable of coordinating multiple GPU/AI accelerator cards, well‑suited for generative‑AI inference and high‑throughput AI service deployment.
4. Dual‑Socket Rack‑Mount Server Deployment
Support dual‑socket configuration via four high‑speed 24GT/s UPI links. Widely adopted for building flagship dual‑socket rack‑mount servers for scenarios requiring extreme multi‑thread scalability.
5. High‑End Server Refurbishment & Bulk Enterprise Hardware Replacement
Flagship Xeon 6 generation CPU with stable FCLGA4710 platform compatibility. Frequently used for premium refurbished server reconstruction and enterprise high‑end equipment upgrade projects.