Overview: This workbook provides granular unit shipment forecasts and supply chain market share data from 2024 through 2028, supporting our latest research note and PR: "HBM Demand for AI Server Compute ASICs to Grow 35x by 2028."

Key Insights & Datasets:

Hyperscaler Shipment Forecasts: Detailed unit breakdowns for top CSPs (including Google, AWS, Microsoft, and Meta) as the market accelerates toward vertical integration.

Vendor Market Share Analysis: Evolution of the design partner landscape, tracking the competitive dynamics between incumbents (Broadcom, Marvell) and challengers (MediaTek, Alchip).

Supply Chain Trend: Analysis of the structural pivot from training-focused silicon to cost-optimized inference solutions and the emerging impact of dual-sourcing strategies.

The workbook includes a detailed AI Server Compute ASIC Shipment Outlook, fully segmented by the following dimensions:

• CSP: (Google, Microsoft, AWS, Meta, xAI, OpenAI, ByteDance, etc.)
• ASIC Model: (TPU, Trainium, Maia, MTIA, etc.)
• Design Partner: (Broadcom, Marvell, Alchip, MediaTek, GUC, etc.)
• Foundry: (e.g., TSMC)
• Process Node: (5nm, 3nm, 2nm, etc.)
• Packaging Type: (CoWoS, EMIB, SoIC, etc.)
• Memory Configuration: (HBM Type)

Table of contents:

• Assumptions and Analysis
• Definitions
• Summary
• Pivot Table
• Flatfile

Published Date: 12 March 2026