ACCESS THE REPORT

AI Processors and Accelerators Used in Data Center Market for AI Applications

1H 2024, Global Market Sizing and Forecast Report, 2023 - 2028

Futurum Intelligence has released a comprehensive analysis of the global market and five-year forecast focused on AI processor and accelerator chipsets used in data centers for AI training and inference applications. 

The 30-page report includes:

  • Detailed 2023 revenue and growth through 2028 by product type
  • Deployment type by year
  • Consumption by buyer persona
  • 11 use cases, sized by year
  • 17 vertical markets, sized by year

The report also includes the market dynamics and critical assumptions we believe will reflect the purchasing behaviors of enterprise buyers, hyperscalers, and data center providers as they manage the increasing and varied compute requirements for AI applications. 

The report is available to Futurum Intelligence subscribers and available in our platform. If you are not a current subscriber, please complete the form below to request access.

ABOUT THE REPORT

This 30-page report provides fresh data and deep insights into 2023 market sizing and a five-year forecast (2024 - 2028) of vendor revenue for processor and accelerator chipsets used in data centers for AI training and inference applications. 

The product categories Futurum studied were CPUs, GPUs, XPUs, and Public Cloud AI Accelerators:

  • Central Processing Units (CPUs) – The CPU is a general-purpose processor capable of handling many computing tasks via multiple cores. While a CPU can support some AI workloads, its sequential versus parallel processing architecture may not be efficient for complex deep learning tasks in data centers that demand high throughput. Examples: NVIDIA Grace, AMD EPYC, Intel XEON, IBM Z Telum and Power10.
  • Graphics Processing Units (GPUs) – GPUs are powerful tools for AI workloads within data centers. Their high parallel architecture supports processing multiple tasks simultaneously for both training and running deep learning models. Examples: NVIDIA H100, AMD Instinct, Intel MAX & FLEX Series
  • XPUs – An XPU is a collective term used for a dedicated accelerator used in artificial intelligence applications. Examples: Intel Gaudi, IBM Telum on-chip accelerator, and Groq LPU.
  • Public Cloud AI Accelerators – These custom accelerators are specifically built for or by public cloud providers for use in artificial intelligence applications. For this study, they are distinct from XPUs. Examples: Google TPU, AWS Trainium and Inferentia, and Microsoft Maia.

Futurum studied 18 chip vendors for this report and estimated their revenue for each product category. These vendors are:

  • Alibaba
  • AMD
  • AWS
  • Baidu
  • Cerebas Systems
  • Google Cloud
  • Graphcore
  • Groq
  • Huawei
  • IBM
  • Imagine Technologies
  • Intel
  • Microsoft
  • Mythic
  • NVIDIA
  • Qualcomm
  • SambaNova Systems
  • SiFive

Submit the form above to request access to the complete report.

tfg_ai-processing-in-data-centers-08192024

The data in the above infographic is derived from the Futurum Intelligence report, AI Processors and Accelerators Used in Data Center Market for AI Applications: 1H 2024, Global Market Sizing and Forecast Report, 2023 - 2028.