Ampere Computing | ASA Computers

The World's First Cloud Native Processors

Ampere is designing the future of hyperscale cloud and edge computing with the world's first cloud native processor. Built for the cloud with a modern 64-bit ARM server based architecture, Ampere gives customers the freedom to accelerate the delivery of all cloud computing applications. With industry-leading performance, power efficiency and linear scalability, Ampere processors are tailored for the continued growth of cloud and edge computing. Ampere and ASA Computers are both members of the AI Platform Alliance to enable deployment for purpose-built AI solutions with easy adoption and seamless usage across all spectrum of AI use cases.

MODERN DATA CENTER READY

Ampere's AmpereOne^® M is a high-performance, cloud-native processor designed for dense AI compute and modern data center workloads. With up to 192 Arm^® cores, 12-channel DDR5 memory, and PCIe Gen5, it enables high-density AI inference, strong performance per rack, and improved energy efficiency, delivering scalable, power-efficient compute for cloud and enterprise environments.

Products

Ampere Product Overview

Processor	AmpereOne^® M	AmpereOne^®	Ampere^® Altra Max	Ampere^® Altra
Primary Workloads	GenAI, AI inference, memory-intensive compute	Cloud, AI, general compute	High-density cloud compute	Cloud & edge compute
Cores	96 - 192	96 - 192	96 - 128	32 - 80
Core Design	Single-threaded cores	Single-threaded cores	Single-threaded cores	Single-threaded cores
L2 Cache per Core	2 MB	2 MB	1 MB	1 MB
System Level Cache (SLC)	Up to 64 MB	Up to 64 MB	Up to 16 MB	Up to 32 MB
Memory Type	DDR5	DDR5	DDR4	DDR4
Memory Channels	12	8	8	8
Max Memory Capacity	Up to 1.5 TB	Up to 4 TB	Up to 4 TB	Up to 4 TB
PCIe Connectivity	96 x PCIe Gen5	128 x PCIe Gen5	128 x PCIe Gen4	128 x PCIe Gen4
TDP Range	250 - 425 W	200 - 400 W	190 - 250 W	45 - 250 W
Key Differentiator	Maximum memory bandwidth for AI	High core density & performance	Higher frequency cloud SKUs	Power-efficient scale-out compute

WHY AMPERE?

Linear Scalability

With leading power/core, and multi-socket support, Ampere provides linear scalability to maximize the number of cores per rack, unparalleled in the industry.

Power Efficiency

Organizations can reduce data center operational expenses with Ampere by decreasing energy consumption, space requirements, optimizing rack allocation, and reducing environmental impact.

Predictable High Performance

Enable to run consistently at high frequencies and are built with large low-latency private caches to deliver high performance even under maximum load conditions.

Ampere Optimized AI Frameworks

With Ampere, customers can achieve superior performance for AI workloads by integrating optimized inference layers into common AI frameworks.

Framework Integration Layer

Provides full compatibility with popular developer frameworks. Software works with the trained networks "as is". No conversions are required.

Model Optimization Layer

Implements techniques such as structural network enhancements, changes to the processing order for efficiency, and data flow optimizations, without accuracy degradation.

Hardware Acceleration Layer

Includes a "just-in-time", optimization compiler that utilizes a small number of Microkernels optimized for Ampere processors. This approach allows the inference engine to deliver high-performance on all frameworks.

Ampere Model Library

AML is a collection of optimized AI models pretrained on standard datasets. The library contains scripts running the most common AI tasks. The models are available for Ampere customers to quickly and seamlessly build into their applications.