PUBLISHER: IDC | PRODUCT CODE: 1480984
PUBLISHER: IDC | PRODUCT CODE: 1480984
IDC's Worldwide Accelerated Computing Platforms and Systems Taxonomy, 2024
This IDC study is a refresh of IDC's worldwide accelerated computing taxonomy last published in 2019. It builds on the foundational segmentations and adds a more comprehensive approach to mapping use cases to the accelerator type."We are seeing the expansion of accelerated computing due to the accelerated adoption of AI, big data and analytics, and modeling and simulation in the datacenter," said Kuba Stolarski, research vice president, IDC's Worldwide Infrastructure research. "Accelerated computing is fundamentally shifting how servers are designed and procured, making technologies like liquid cooling more mainstream."
IDC's Worldwide Accelerated Computing Platforms and Systems Taxonomy
Accelerated Computing Platforms and Systems Taxonomy Changes for 2024
Taxonomy Overview
Hierarchical Mapping: Accelerated Computing Platforms and Systems Primary/Secondary Markets
Definitions
- Accelerator Characteristics
- Accelerator Design Type, Architecture, and Form Factor
- Design Type
- Architecture
- Graphics Processing Unit (GPU)
- Field-Programmable Gate Array (FPGA)
- Application-Specific Integrated Circuit
- Hybrid/Composite
- Form Factor
- Discrete
- Integrated Onboard
- Integrated on Chip (Embedded)
- Open Compute Project Open Domain-Specific Architecture
- Computational Characteristics
- Memory
- 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
- Core Count
- Floating-Point Performance
- Dynamic Performance and Precision
- Computational Performance
- Comparing Accelerator Performance
- Memory
- Interconnect and Access Software
- Accelerator/Platform Interconnect
- Peripheral Component Interconnect Express
- Computer Express Link
- Other Emerging Interconnect Standards
- Access Software
- Accelerator-Specific SDKs and APIs
- Open Standards-Based SDKs and APIs - Open Computing Language
- Other SDKs and APIs
- Virtualization and Partitioning
- Accelerator/Platform Interconnect
- Power and Cooling
- Thermal Design Power
- Cooling
- Accelerator Design Type, Architecture, and Form Factor
- Computing Platform Characteristics
- Form Factors, Attach Rate, and Attach Type
- Form Factors
- Small Form Factor Racks and Non-Rack Form Factors
- Medium-Sized Rack-Based Form Factors
- Large Standalone Rack-Scale Form Factors
- Attach Rate
- Attach Type
- Form Factors
- Form Factors, Attach Rate, and Attach Type
- Computing System Characteristics
- System Type and Convergence
- System Type
- Convergence
- Scaling and Networking
- Scaling
- Level 1 (Intra-Silicon)
- Level 2 (Intraplatform)
- Level 3 (Intrasystem)
- Level 4 (Intersystem)
- Networking
- Scaling
- System Type and Convergence
- Procurement, Installation, and Deployment
- Shipment Type
- Factory Installed
- Reseller or SI Installed
- End User Installed
- Buyer Type
- Hyperscalers
- Service Provider
- Non-Service Provider
- Deployment Location
- Deployment Type
- Dedicated (Private) Cloud
- Shared (Public) Cloud
- Noncloud
- Edge
- Shipment Type
- Use Cases, Workloads, and Offload Type
- Use Cases
- Industry-Specific Use Cases
- Artificial Intelligence
- Modeling and Simulation (aka High-Performance Computing)
- Data Analytics Use Cases
- Other Application Use Cases
- Industry-Agnostic Use Cases
- Datacenter Hyper-Customization
- IT Infrastructure
- Industry-Specific Use Cases
- Workloads
- Offload Type
- (General) Application-Offload
- (Embedded) Function-Offload
- Use Case to Server Workloads Mapping
- Use Cases
Related Markets
Learn More
- Related Research
- Synopsis
