By using this site, you agree to the IDC Privacy Policy

Table of Contents

IDC's Worldwide Accelerated Computing Platforms Taxonomy

Figure: Accelerated Computing Platforms Primary Segments

Accelerated Computing Platforms Taxonomy Changes for 2019

Taxonomy Overview

Definitions

Deployment Location

Traditional IT and Private Cloud — On- and Off-Premises

Private Cloud Deployments

Public Cloud — On- and Off-Premises

Accelerated Public Cloud Instances

Accelerator-Based Clouds

Public Cloud On-Premises (Service Delivery Edge)

Edge (IT, OT, and CT)

Accelerated Computing Platform Architecture

Integrated Platforms

Server Vendor (OEM) Integrated

VAR or SI Integrated

Discrete Unit (Add-On Card)

Accelerator Design

Accelerator Architecture

Figure: Accelerator Design Secondary/Subsegments

Graphics Processing Unit

Advanced Micro Devices
NVIDIA
Intel

Manycore Microprocessor

Coprocessor

Field-Programmable Gate Array

Application-Specific Integrated Circuit

Hybrid

Accelerator/Platform Interconnect

Peripheral Component Interconnect Express

NVLink (NVIDIA)

Infinity Fabric (AMD)

Computer Express Link

Other Emerging Interconnect Standards

Accelerator Form Factor

Discrete (Peripheral Component Interconnect Express Based)

Discrete (Peripheral Component Interconnect Mezzanine Based)

Discrete (Mobile Peripheral Interconnect Express Module Based)

Integrated on Board

Integrated System on a Chip

Open Compute Project Open Domain-Specific Architecture

FPGA-Specific Form Factors

Accelerator Memory Type

Synchronous Graphics Random Access Memory Double Data Rate

High-Bandwidth Memory

Embedded Dynamic Random Access Memory

Synchronous Dynamic Random Access Memory

Resistive Random Access Memory

Accelerator Thermal Design Power

Accelerator Core Count

Floating-Point Performance

Dynamic Range and Precision

Computational Performance

Comparing Accelerator Performance

Cooling

Active Cooling

Passive Cooling

Accelerated Computing Software

Accelerator-Specific SDKs and APIs

Graphics Virtualization

Open Standards–Based SDKs and APIs

Open Computing Language

Other SDKs and APIs

Workloads and Use Case Segmentation

Server Workloads

Figure: Accelerated Compute Use Cases Mapping

Embedded Workloads

Industry-Specific Use Cases

Artificial Intelligence Use Cases

Traditional Accelerated Server Use Cases

Data Analytics Use Cases

Additional Use Cases

Datacenter Hyper-Customization

Use Case to Server Workloads Mapping

Figure: Apportioning Accelerated Compute from Use Cases to Workloads

Learn More

Related Research

Synopsis

Content


Related Links