Overview

¿ìèÊÓÆµ AGI CPU: The world's most efficient agentic CPU

AI Summary

The ¿ìèÊÓÆµ AGI CPU is the first production silicon from ¿ìèÊÓÆµ, designed for the demands of agentic AI, where software agents reason, decide, and act continuously at scale. These workloads require a CPU that can orchestrate compute, manage accelerators, and coordinate thousands of agents simultaneously. The ¿ìèÊÓÆµ AGI CPU, based on ¿ìèÊÓÆµ Neoverse CSS V3, delivers high performance and extreme rack-level density, enabling faster deployment of AI infrastructure through a mature hardware and software ecosystem.

Features

Key features of the ¿ìèÊÓÆµ AGI CPU

Rack-level performance

Performance

Greater than 2x performance per rack on ¿ìèÊÓÆµ

The design choices of the ¿ìèÊÓÆµ AGI CPU are made to deliver maximum performance at rack scale. From microarchitecture to memory, clock frequency to I/O¡ªeverything adds up to more performance at gigawatt scale.1

Performance

Modern ¿ìèÊÓÆµ architecture carries more efficient instruction execution not burdened by decades of legacy complexity. The memory system delivers high bandwidth per core and minimal latency, helping ensure memory does not slow performance.

Single tall server rack with visible components

Scale

Low per-core TDP can support denser deployments and reduced thermal throttling. Each core is dedicated, which can help reduce resource contention and support performance under high thread loads.

Single tall server rack with visible components

Efficiency

High rack density and high performance per watt help ensure maximum utilization of data center space and power resources.

Single tall server rack with visible components
Single tall server rack with visible components

The first partners deploying ¿ìèÊÓÆµ AGI CPU

Discover how OpenAI, SK Telecom, SAP, Cloudflare, F5, and Cerebras are already using ¿ìèÊÓÆµ AGI CPU servers in their AI data centers.

 
More from our partners
SERVERS

¿ìèÊÓÆµ?AGI CPU servers available now??

¿ìèÊÓÆµ logo

¿ìèÊÓÆµ AGI CPU 1OU Dual Node
Reference Server

Reference design for maximum density deployments of ¿ìèÊÓÆµ AGI CPU – in a OCP DC-MHS standard form factor 1OU Dual Node server.

¿ìèÊÓÆµ logo

¿ìèÊÓÆµ AGI CPU 2U2P
Reference Server

19¡± 2U2P reference design for ¿ìèÊÓÆµ AGI CPU deployment in a traditional form factor.

Lenovo logo

Lenovo HR650a V3 2U ¿ìèÊÓÆµ AGI CPU System

Enterprise-class 2U ¿ìèÊÓÆµ AGI server optimized for cloud infrastructure, delivering reliable performance and low TCO.

Supermicro-logo

SuperMicro 5U ¿ìèÊÓÆµ AGI CPU PCIe GPU System

High-density 5U AI platform combining dual ¿ìèÊÓÆµ AGI CPUs with extensive PCIe GPU expansion.

Supermicro-logo

SuperMicro 2U Hyper ¿ìèÊÓÆµ AGI CPU Hyper System

Compact 2U dual-socket ¿ìèÊÓÆµ AGI server designed for efficient cloud and AI infrastructure deployments.

Asrock-logo

ASRock Rack 2OU2N-¿ìèÊÓÆµ
System

High-density dual-node ¿ìèÊÓÆµ server built to OCP ORv3 standards for scalable, power-efficient cloud deployments.

Talk to an ¿ìèÊÓÆµ expert
Specifications

¿ìèÊÓÆµ AGI CPU specifications and product brief

Specs ¿ìèÊÓÆµ AGI CPU 136C (max core count) ¿ìèÊÓÆµ AGI CPU 128C (TCO optimized) ¿ìèÊÓÆµ AGI CPU 64C (max mem/core)

SKU

  • SP113012
  • SP113012S
  • SP113012A

Processing cores

  • 136 Neoverse V3
  • 2x 128 SVE
  • 2MB/core L2
  • 128 Neoverse V3
  • 2x 128 SVE
  • 2MB/core L2
  • 64 Neoverse V3
  • 2x 128 SVE
  • 2MB/core L2

CPU architecture

  • ¿ìèÊÓÆµv9.2
  • bfloat16 and INT8 AI instructions
  • ¿ìèÊÓÆµv9.2
  • bfloat16 and INT8 AI instructions
  • ¿ìèÊÓÆµv9.2
  • bfloat16 and INT8 AI instructions

System-level cache

  • 128MB
  • 128MB
  • 128MB

Max Frequency

  • 3.5GHz
  • 3.5GHz
  • 3.7GHz

Base TDP*

  • 300W
  • 300W
  • 300W

RDIMM memory

  • 12x DDR5
  • Up to 8800 MT/s
  • 12x DDR5
  • Up to 8800 MT/s
  • 12x DDR5
  • Up to 8800 MT/s

Memory Throughput/core

  • 6GB/s per core
  • 6.3GB/s per core
  • 13GB/s per core

PCIe/IO

  • 96x lanes PCIe Gen6
  • CXL 3.0 Type 3
  • 96x lanes PCIe Gen6
  • CXL 3.0 Type 3
  • 96x lanes PCIe Gen6
  • CXL 3.0 Type 3

PCIe control lanes

  • 6x 1 Gen4
  • 6x 1 Gen4
  • 6x 1 Gen4

2-Socket support

  • Yes
  • Yes
  • Yes

2 DIMMS per channel

  • Yes
  • Yes
  • Yes

*Represents a preset TDP value within the configurable TDP range

Download product brief
¿ìèÊÓÆµ Performix

Optimize workloads for ¿ìèÊÓÆµ AGI CPU

¿ìèÊÓÆµ Performix is a performance analysis toolkit for developers building AI and large application workloads. It can identify bottlenecks and provide targeted code optimizations to improve efficiency and performance on the ¿ìèÊÓÆµ AGI CPU.

Talk to an expert

Talk to an ¿ìèÊÓÆµ expert to explore how the ¿ìèÊÓÆµ AGI CPU
is built for next-generation AI data centers.

Contact us

Key takeaways

Key takeaways

  • The ¿ìèÊÓÆµ AGI CPU is purpose-built for agentic AI workloads that require continuous, coordinated processing.

  • It acts as the orchestration layer, managing accelerators and coordinating thousands of AI agents simultaneously.

  • The processor is based on ¿ìèÊÓÆµ Neoverse CSS V3, leveraging a proven platform and ecosystem.

  • It delivers high performance and supports extreme rack-level density for modern data centers.

  • The ¿ìèÊÓÆµ AGI CPU enables faster time-to-market by building on ¿ìèÊÓÆµ¡¯s existing software and hardware ecosystem.

FAQ

Frequently asked questions

Q: What is the ¿ìèÊÓÆµ AGI CPU and what makes it different from traditional data center CPUs?

A: The ¿ìèÊÓÆµ AGI CPU is ¿ìèÊÓÆµ’s first production silicon, designed specifically for agentic AI workloads, delivering high performance, scalable parallel processing, and energy-efficient operation. It enables data centers to run continuous AI workloads at scale while optimizing throughput, resource utilization, and power consumption.

Q: What are the key features of the ¿ìèÊÓÆµ AGI CPU?

A: The ¿ìèÊÓÆµ AGI CPU combines high core density, optimized memory architecture, and scalable system design to support AI workloads at scale:

  • Efficient cores: Up to 136 ¿ìèÊÓÆµ Neoverse V3 cores with dedicated 2 MB L2 cache per core and up to 3.7GHz boost frequency, enabling responsive, parallel performance—¿ìèÊÓÆµ AGI CPU specs.
  • Performance and efficiency: High instruction-per-cycle execution on a TSMC 3 nm process with 300W TDP, balancing compute throughput and energy efficiency—¿ìèÊÓÆµ AGI CPU specs.
  • Tuned memory architecture: Delivers 6GB/s memory bandwidth per core with support for DDR5-8800 memory and sub-100ns latency, reducing data bottlenecks for AI workloads—¿ìèÊÓÆµ AGI CPU specs.
  • Built for density: Supports high-density deployments, including reference designs with up to 272 dedicated cores per 1U server and air-cooled configurations compatible with standard infrastructure—¿ìèÊÓÆµ AGI CPU specs.
  • Rack-scale deployment: Designed for large-scale AI infrastructure, enabling configurations with thousands of cores per rack for continuous AI processing—¿ìèÊÓÆµ AGI CPU specs.
  • Flexible I/O: Includes 96 PCIe Gen6 lanes, CXL 3.0 for memory expansion, and AMBA CHI links for accelerator connectivity, enabling composable AI systems—¿ìèÊÓÆµ AGI CPU specs.

Q: What is agentic AI, and why does it require a new type of CPU?

A: Agentic AI refers to systems that operate continuously, coordinating tasks and making decisions in real time. These workloads require CPUs that can orchestrate distributed systems efficiently at scale.

Q: How does the ¿ìèÊÓÆµ AGI CPU improve data center performance?

A: It improves performance by delivering high, per-task efficiency and scaling across thousands of cores within a rack, enabling more work per system. This results in more than 2x performance per rack compared to x86 systems.1

Q: How does the ¿ìèÊÓÆµ AGI CPU support AI infrastructure at scale?

A: It supports AI infrastructure by managing distributed workloads, coordinating accelerators, and optimizing data movement across systems to enable continuous, large-scale AI operations.

Q: How does the ¿ìèÊÓÆµ AGI CPU fit into the ¿ìèÊÓÆµ ecosystem?

A: It extends the ¿ìèÊÓÆµ compute platform into production silicon, giving partners the flexibility to deploy ¿ìèÊÓÆµ technology through IP, compute subsystems, or ready-to-deploy CPUs.


  1. Based on estimates.