Ampere Computing

What Ampere Computing does

Ampere Computing is a fabless semiconductor company founded in 2017 and headquartered at 4655 Great America Parkway, Suite 601, Santa Clara, California. The company designs high-performance, energy-efficient Arm-based processors for cloud computing, AI inference, and edge workloads. Its product portfolio is anchored by three processor families: the AmpereOne platform (delivering the best compute efficiency and container/VM density for traditional cloud and legacy AI workloads), the AmpereOne M platform (a 192-core processor targeting dense AI inference with multi-modal and agentic AI-based LLM services at the lowest claimed cost per inference session), and the Ampere Altra platform (high-performance, power-efficient processors widely adopted in telecom, networking, autonomous vehicles, and edge AI applications). Supporting offerings include the Ampere AI Solutions platform for sustainable inference, the Ampere Developer Center with reference architectures for Spark, Kubernetes, and AI workloads, and the Ampere Ready Software catalog validating 130+ packages on its processors.

The company's business model centers on hardware sales of its Arm-based processors to cloud service providers, system integrators, and enterprise customers, distributed via a multi-channel model that includes direct hyperscaler marketplace listings (Oracle Cloud Infrastructure A4 instances, Microsoft Azure Dpsv5/Epsv5 VMs, Google Cloud Tau T2A VMs), OEM partnerships (Supermicro, GIGABYTE, HPE, Foxconn Industrial Internet, Wiwynn), and global distributors (Arrow, Avnet, Edom, Newegg, TD SYNNEX). Pricing is quote-based and not publicly disclosed. Customer segments span cloud service providers and hyperscalers (primary), enterprise data centers, edge and embedded computing, media services, and research/HPC. Named end-customer case studies include Oracle Cloud Infrastructure, Microsoft Azure, Google Cloud, ZEISS, 8x8, University of Miami, and Milwaukee School of Engineering, with documented cost reductions of 60-80% and performance gains up to 20x versus x86-based solutions.

In November 2025, SoftBank Group completed its $6.5 billion acquisition of Ampere via subsidiary Silver Bands 6 following FTC clearance, making Ampere a wholly owned subsidiary within SoftBank's AI Computing Segment alongside Arm Holdings and Graphcore. Ampere was not publicly traded prior to acquisition and has never disclosed revenue. Following acquisition, the company continues to operate as a semiconductor design house, with a recent European expansion (March 2026) bringing AmpereOne-powered instances to Oracle Cloud in London and Frankfurt, plus regional providers including Scaleway, Hetzner, Glesys, C41.ch, and CloudSigma.

Name

Ampere Computing

Legal name

Ampere Computing LLC

Website

https://amperecomputing.com

Company type

Private

Founded year

2017

Operating status

Operating

Headcount range

1,001–5,000 employees

Short description

Ampere Computing is a fabless semiconductor company that designs energy-efficient Arm-based processors for servers, cloud computing, and AI inference workloads, serving hyperscalers, enterprises, and edge customers. Acquired by SoftBank for $6.5 billion in November 2025.

Ownership category

akta.pro rank

Product category

Server Processors

NAICS

Semiconductor and Other Electronic Component Manufacturing (3344)

SIC

Electronic Computers (3571), Electronic Components, Nec (3679)

akta.pro primary industry

Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC) (HDAHABAC)

akta.pro secondary industries

Application Processors & Microprocessors (MPUs/CPUs) (HDAHAJAB), AI/Neural Processing Units (NPUs) & ML Accelerators (HDAHAJAF), Hyperscale/OEM & Open Compute (OCP) Server Platforms (HDACABAI)

Keywords

• Arm-based server processors
• Cloud native processors
• AI inference chips
• Data center CPUs
• Semiconductor design

Headquarters

HQ city

Santa Clara

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Hardware or Manufacturing

Cost components

Personnel, Technology or R&D, Operations, Marketing or Sales, Supply Chain

Revenue model

1. Processor Sales: Ampere generates revenue through the sale of its Arm-based processors (Ampere Altra, AmpereOne families) to cloud service providers, system integrators, and enterprise customers.

Go-to-market motion1 record

Distribution channels7 records

Marketing channels6 records

Core offering

Ampere Computing designs and sells Arm-based server-class processors (Ampere Altra, AmpereOne, and AmpereOne M families) optimized for cloud native workloads, dense AI inference, and edge applications. The fabless semiconductor company provides high core-count CPUs with single-threaded performance, energy efficiency, and a validated software ecosystem for hyperscalers, enterprises, and OEMs.

Product overview

Ampere Computing is a semiconductor design company producing Arm-based cloud native processors for AI, cloud, and edge computing. Its product portfolio is organized around two main processor families — AmpereOne and AmpereOne M for advanced AI and dense cloud workloads, and Ampere Altra for power-efficient general-purpose and edge computing — supplemented by the Ampere AI Solutions platform for sustainable inference, the Ampere Developer Center for tooling and reference architectures, and the Ampere Partner Ecosystem encompassing major cloud providers (Oracle, Microsoft Azure, Google Cloud), system integrators (Supermicro, GIGABYTE, HPE), and a validated Ampere Ready Software catalog. The company was acquired by SoftBank Group in late 2025 for $6.5 billion and positions its processors as efficient alternatives to GPU-based AI compute.

Differentiator

Problem solved

Functional benefit

Products and services

• AmpereOne Server-class processor platform offering the best compute efficiency and container/VM density for both traditional workloads and legacy AI models. Delivers high core counts with single-threaded performance and scalable efficiency from cloud to edge.
• AmpereOne M Advanced compute processor platform delivering dense AI performance for multi-modal and agentic AI-based LLM services with the lowest cost per inference session available. Features up to 192 cores designed for dense AI inference environments.
• Ampere Altra Extremely efficient, high-performance cloud native processor for power-constrained racks and sensitive environments. Widely adopted in telecom, networking, autonomous vehicles and edge AI applications, offering scalable and efficient performance for traditional and AI-enhanced services.
• Ampere AI Solutions Sustainable AI inference platform combining Ampere processors with partner accelerator technologies. Delivers AI compute with a fraction of the power and resources required by GPU-based solutions, supporting TensorFlow, PyTorch, and ONNX frameworks for cloud native AI workloads.

Quantifiable outcome

• Up to 10x reduction in drug discovery simulation runtime
• +4 more outcomes

Named customers10 records

Segments5 records

Ideal customer profiles5 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Integration6 records

AI capability12 records

Feature6 records

Partnerships

Ten partnerships are on record, tiered core and minor.

• SupermicrocoreTechnology or IntegrationSupermicro's MegaDC series servers powered by AmpereOne and Ampere Altra families excel in AI inference tasks and advanced data management at scale.
• Oracle Cloud InfrastructurecoreTechnology or IntegrationOracle and Ampere offer predictable performance, near-linear scaling, and secure architecture with top-tier price-performance through Ampere-based OCI instances.
• Microsoft AzurecoreTechnology or IntegrationMicrosoft Azure offers Dpsv5 (general-purpose) and Epsv5 (memory-optimized) virtual machines featuring Ampere Altra processors.
• Google CloudcoreTechnology or IntegrationGoogle Cloud's Tau T2A VMs powered by Ampere Altra processors provide versatile solutions ideal for scale-out workloads.
• Hewlett Packard EnterprisecoreTechnology or IntegrationHPE's ProLiant RL300 Gen11 featuring Ampere processors offers predictable frequency scaling, higher data center density, and power efficiency.
• GIGABYTEcoreTechnology or IntegrationGIGABYTE's servers powered by AmpereOne and Ampere Altra processors deliver performance and efficiency for data centers and HPC environments.
• NETINTcoreTechnology or IntegrationNETINT collaborated with Supermicro and Ampere to build a video transcoding server combining NETINT's Quadra VPUs with Ampere Altra processors, achieving 80% cost reduction and 20x throughput increase.
• Arrow ElectronicsminorChannel Partner/ Reseller/ DistributorArrow Electronics distributes Ampere products globally.
• AvnetminorChannel Partner/ Reseller/ DistributorAvnet distributes Ampere products as part of its electronics distribution network.
• TD SYNNEXminorChannel Partner/ Reseller/ DistributorTD SYNNEX distributes Ampere products as part of its technology distribution portfolio.

Scale indicators7 records

Recent moves8 records

Expansion highlights5 records

Direct peers

• Marvell Technology: Fabless designer of custom Arm-based data center silicon (including server CPUs, DPUs, and NICs) sold to hyperscalers. Marvell operates the same fabless, custom-and-merchant Arm server silicon business model targeting the same hyperscaler and cloud customers as Ampere.

Broad incumbents

• Intel Corporation: Incumbent x86 data center CPU supplier whose Xeon family dominates installed enterprise server bases. Intel is the primary incumbent Ampere targets displacing in mainstream enterprise data centers, competing on performance per watt and total cost of ownership.
• AMD: Challenger x86 data center CPU supplier whose EPYC processors compete directly with Ampere for cloud-native and AI inference workloads at hyperscalers. AMD is the most relevant x86 alternative Ampere benchmarks against on performance, efficiency, and price.
• NVIDIA: Dominant supplier of GPUs and AI accelerators, and the reference platform for training and much of large-model inference. NVIDIA is the principal incumbent whose AI compute stack Ampere's CPU-centric and hybrid inference positioning is built to complement and partially displace.
• Broadcom: Fabless designer of custom ASICs and networking silicon for hyperscalers, including merchant and custom Arm-based server solutions. Broadcom competes with Ampere for hyperscaler design wins and operates a similar fabless model targeting the same cloud customers.

Emerging players

• SiPearl: European fabless designer of Arm-based high-performance processors for supercomputing and HPC, with EU sovereign-cloud backing. SiPearl is an emerging Arm server CPU peer with overlapping architectural approach and a focus on regulated European workloads similar to Ampere's European expansion.
• Rivos: Early-stage fabless startup designing Arm-based server CPUs targeting data analytics and AI workloads. Rivos is an emerging direct peer pursuing the same Arm-based server CPU market as Ampere, competing for venture capital, talent, and design wins.
• Qualcomm: Arm IP holder pursuing entry into data center and cloud server CPUs following its Nuvia acquisition. Qualcomm represents a credible emerging Arm-based server CPU competitor leveraging deep Arm IP and mobile ecosystem expertise.
• Graphcore: AI accelerator company now part of SoftBank's AI Computing Segment alongside Ampere and Arm. Graphcore is a sibling company under SoftBank and an AI compute peer targeting overlapping inference and training workloads in cloud and enterprise deployments.
• Cerebras Systems: AI compute company selling wafer-scale accelerator systems for training and inference in data centers. Cerebras is an emerging AI infrastructure peer pursuing a different compute substrate but addressing the same large-model and inference workloads that Ampere's CPU-centric stack targets.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights6 records

Customer concentration

• Xx.com/AmpereComputing
• GitHubgithub.com/AmpereComputing
• YouTubeyoutube.com/channel/UC_r7FyrSOv63sm6f3K1M24A
• Facebookfacebook.com/AmpereComputing
• LinkedInlinkedin.com/company/amperecomp

Revenue estimate

Valuation estimate

Number of profiles

Profiles8 records

Funding overview

Funding rounds7 records

Investors3 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A1 record

Investments1 record

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Ampere Computing