semianalysis.com

SemiAnalysis is an independent research firm covering the AI and semiconductor supply chain, read by 318,000+ subscribers. Beyond its Substack newsletter, it offers four institutional research products — an Accelerator & HBM Model, a Datacenter Industry Model, a Tokenomics Model, and open-source tooling and benchmarks — plus consulting and API access. The models are aimed at institutional users who need bottoms-up forecasts and supply-chain data rather than news commentary.

The Four Research Products

Accelerator & HBM Model

A bottoms-up demand forecast for AI accelerators and high-bandwidth memory, covering Nvidia, AMD, Google, AWS, Meta, Microsoft, Intel, and every major Chinese hyperscaler.

• Granularity: SKU-level shipments and ASPs, with customer-level shipments and installed base for 60+ accounts.
• HBM detail: tracked down to vendor, generation, layer count, and price — the same depth as compute.
• Supply-chain link: shipments are built up from equipment requirements, wafer starts, and packaging orders, so revenue for supply-chain companies can be estimated directly from production.
• Coverage: forecasts to 2030, history back to 2021.

This is the product to look at if your question is "how many accelerators ship to whom, and what does that imply for memory and upstream suppliers."

Datacenter Industry Model

A global datacenter capacity tracker built around structure-level megawatt modeling.

• Scope: 6,500+ structures, including hyperscaler buildouts and colocation markets, with per-structure power draw, lease structures, and planned versus operational capacity.
• Companies: 200+ covered, with the end customer identified behind each megawatt.
• Demand side: accelerator shipments and installed base by operator, power drawn per deployment, and a supply-demand reconciliation showing which operators and regions run short of power, and when.
• Delivery tracking: promised versus delivered MW, tracked quarterly, with regional site strategy.

Use this when the question is about power constraints, buildout pacing, or where capacity is actually landing.

Tokenomics Model

A unit-economics engine for AI serving: bottoms-up token throughput, ROI on AI spend, and translation of demand-side revenue into future accelerator demand by hyperscaler. It connects inference economics back to the hardware forecast — useful if you need to test whether serving revenue can justify projected compute purchases.

OSS Tooling & Benchmarks

Open-source tooling and benchmarks published under a transparency-oriented approach. These are the lowest-commitment entry point: you can inspect methodology and outputs before engaging with the commercial models.

Choosing Between Them

Access and Practical Notes

The site lists a newsletter subscription link, institutional login, and "Talk to Sales" / "View model" paths for each product, along with consulting and API access offerings. The available material does not state prices or whether any tier is free, so treat access terms as something to confirm directly — the newsletter is the public-facing entry point, while the models appear to be positioned as institutional offerings.

A reasonable sequence: read the newsletter and OSS benchmarks to judge research quality, then request model access or a sales conversation for the specific product matching your question.

SemiAnalysis institutional models and advisory are aimed at organizations that need bottoms-up, SKU- and site-level visibility into the AI and semiconductor supply chain — typically chipmakers, hyperscalers, equipment and memory vendors, and institutional investors building supply, capacity, or demand forecasts. If your work requires customer-level shipment estimates, per-structure datacenter power modeling, or inference unit economics rather than headline market summaries, the institutional offerings are the relevant tier. Casual readers and individual subscribers are served by the newsletter instead.

What the institutional offerings cover

SemiAnalysis describes itself as independent research on the AI and semiconductor supply chain, read by 318,000+ subscribers, with fourteen institutional models, open benchmarks, and advisory. The page evidence lists four model families plus consulting and API access:

The page also lists Consulting ("Engage the best") and API Access for programmatic semiconductor data.

Who each model fits

Accelerator & HBM Model

Best for supply chain and investment analysts who need to connect upstream capacity to downstream deployment. Because shipments are built up from equipment requirements, wafer starts, and packaging orders, the page notes that revenue for companies across the supply chain can be estimated directly from production. Consider this if you need customer-level shipments and installed base rather than aggregate market totals, or if HBM pricing by generation and layer count matters to your thesis.

Datacenter Industry Model

Best for teams assessing power constraints and site strategy. The structure-level database (6,500+ structures, planned vs. operational MW, quarterly delivery tracking) plus a supply-demand reconciliation showing which operators and regions run short of power — and when — suits infrastructure planning, colocation market analysis, and regional deployment questions.

Tokenomics Model

Best for those evaluating AI serving economics: token throughput, ROI on AI spend, and how revenue translates into future accelerator demand. Useful when your question is demand-side (what serving economics imply for compute purchases) rather than supply-side.

Consulting and API Access

Consulting fits organizations that want direct engagement with the analysts rather than model access alone. API Access fits teams that want to pull semiconductor data programmatically into their own systems.

How to evaluate fit and get access

1. Match your question to a model. If you need SKU-level accelerator and HBM forecasts with customer detail, start with the Accelerator & HBM Model. If your question is power, sites, or regional capacity, start with the Datacenter Industry Model. If it is inference economics, start with Tokenomics.
2. Check the coverage window against your needs. Accelerator & HBM coverage runs from 2021 history to 2030 forecasts; datacenter buildouts are covered through 2030. Confirm the Tokenomics horizon directly, since the page does not state it.
3. Use the page's own paths. Each model page offers "View model" and "Talk to Sales." The newsletter has a separate subscribe link (newsletter.semianalysis.com/subscribe), and there is an institutional login for existing clients.
4. Ask about access terms before committing. The page does not publish institutional pricing or contract terms; those come through sales. The newsletter subscription is a distinct product from institutional model access — do not assume one implies the other.

Practical caveats

• Detail level is the selling point, and the workload. Customer-level shipments for 60+ accounts and per-structure power for 6,500+ structures are valuable precisely because they require interpretation. Budget analyst time to use them, not just license cost.
• Forecast horizons are long but not uniform. The page specifies 2030 for accelerators/HBM and datacenter buildouts; Tokenomics horizon is unstated, so verify it if long-range inference economics are central to your use case.
• Independent research positioning. SemiAnalysis presents itself as independent research; if your organization requires specific vendor-relationship disclosures or methodology documentation, request those through sales.
• Start narrow. A single model tied to your primary question is a cleaner trial than a broad institutional package, and the "Talk to Sales" path lets you scope before committing.

If your need is general industry awareness, the newsletter is the lower-commitment entry point. If your need is bottoms-up, account- and structure-level data with multi-year forecasts for supply chain, capacity, or investment work, the institutional models and advisory are the offerings designed for that.

SemiAnalysis offers four access paths: a Substack newsletter, an institutional login, API access for programmatic semiconductor data, and a consulting/sales contact route. Which one fits depends on whether you need free editorial content, organization-wide research models, machine-readable data, or custom advisory work.

Access paths at a glance

Newsletter subscription

The newsletter is the simplest entry point. The site links to a subscribe page at newsletter.semianalysis.com/subscribe, and the homepage opens with an email prompt ("What's your email?"). The publication reports being read by 318,000+ subscribers. Subscription pricing and plan tiers are not specified in the available site material, so check the subscribe page for current terms before assuming any access level.

Institutional login and research models

Organizations can sign in through the "SemiAnalysis Institutional Login" entry on the site. Institutional offerings center on a set of quantitative models, each with its own "View model" and "Talk to Sales" actions:

• Accelerator & HBM Model — bottoms-up demand forecasts for AI accelerators and high-bandwidth memory across Nvidia, AMD, Google, AWS, Meta, Microsoft, Intel, and major Chinese hyperscalers, with customer-level shipments and installed base for 60+ accounts. HBM is tracked by vendor, generation, layer count, and price; forecasts run to 2030 with history back to 2021.
• Datacenter Industry Model — a global capacity tracker covering 6,500+ structures, including hyperscaler buildouts, colocation markets, power constraints, and regional trends through 2030, with per-structure power draw and planned-versus-operational capacity across 200+ companies.
• Tokenomics Model — unit economics for AI serving: bottoms-up token throughput, ROI on AI spend, and translation of revenue into future accelerator demand.

If you are unsure which model matches your question, the "Talk to Sales" link on each model page is the intended first step rather than a direct purchase flow.

API access

The site lists "API Access — Programmatic Semiconductor Data" as a distinct offering. This is the route for pulling semiconductor supply-chain data into your own dashboards or analysis pipelines rather than reading it in report form. The available material does not specify endpoints, authentication, rate limits, or pricing, so treat the API page or a sales conversation as the source of those details.

Consulting and sales contact

For bespoke questions — custom forecasts, advisory engagements, or clarification on model coverage — the Consulting section ("Engage the best") and the repeated "Talk to Sales" links are the contact points. This path makes sense when your need is a specific decision rather than ongoing data access.

Choosing between them

• Reading regular research as an individual: start with the newsletter.
• Needing the full model suite across a team: institutional login, arranged through sales.
• Building data products or automated analysis: API access.
• Needing tailored analysis or a walkthrough of coverage: consulting/sales contact.

One caveat: the site presents these as separate offerings but does not publish a unified price list or state which items are included at which tier. Confirm scope and cost directly through the subscribe page or sales before committing.

SemiAnalysis builds its research from the bottom up: it starts with physical production inputs—equipment requirements, wafer starts, and packaging orders—and works forward to SKU-level shipments, installed base, and revenue. That approach is what distinguishes its models from top-down market sizing, and it is the lens to keep in mind when deciding whether its research fits your needs. The method is best suited to readers who need to connect upstream capacity to downstream deployment—supply chain analysts, infrastructure planners, and investors tracing how fab and packaging constraints translate into available compute.

The Core Mechanism: Build Shipments from Production Inputs

The central design choice is that demand forecasts are not estimated from spending intentions or survey data. Instead, shipments are constructed from what has to be produced to make them possible:

• Equipment requirements feed into capacity estimates.
• Wafer starts translate capacity into die output.
• Packaging orders capture the advanced packaging step that constrains many AI accelerators.

Because each layer is modeled explicitly, revenue for companies across the supply chain can be estimated directly from production rather than inferred from reported sales. The stated coverage runs forecasts to 2030 with history back to 2021.

A concrete example of why this matters: if packaging capacity for high-bandwidth memory is the binding constraint in a given quarter, a top-down demand model will overstate accelerator shipments. A bottoms-up model that tracks packaging orders separately will show the bottleneck before it appears in shipment data—and will also show which upstream suppliers benefit.

Linking Upstream Capacity to Downstream Deployment

The research connects production to where compute actually lands. The Accelerator & HBM Model covers Nvidia, AMD, Google, AWS, Meta, Microsoft, Intel, and every major Chinese hyperscaler, with customer-level shipments and installed base for 60+ accounts. HBM is tracked at the same granularity as compute—down to vendor, generation, layer count, and price.

On the deployment side, the Datacenter Industry Model tracks 6,500+ structures, including hyperscaler buildouts and colocation markets, with per-structure power draw, lease structures, and planned versus operational capacity. More than two hundred companies are covered, with the end customer identified behind each megawatt. A supply-demand reconciliation shows which operators and regions run short of power, and when.

The Tokenomics Model adds a third layer: unit economics of AI serving, including bottoms-up token throughput, ROI on AI spend, and translation of revenue into future accelerator demand.

Independence and Transparency

SemiAnalysis describes itself as independent research on the AI and semiconductor supply chain, read by 318,000+ subscribers, with fourteen institutional models, open benchmarks, and advisory. The transparency claim is backed by open-source tooling and benchmarks published alongside the models—useful if you want to inspect methodology rather than accept conclusions. Note that the site presents these as its own positioning; treat the independence claim as a stated principle and evaluate the models on their published assumptions.

How to Access the Research

Access splits into three tracks, and the right one depends on your use case:

• Newsletter — the Substack publication, positioned as the entry point for individual readers. A subscribe link is available at newsletter.semianalysis.com/subscribe. Pricing and login terms are not specified in the available materials, so check the subscribe page directly before assuming cost or access level.
• Institutional models — the fourteen models above, accessed through an institutional login on the site. Each model page offers "View model" and "Talk to Sales," which indicates a sales-qualified process rather than self-serve signup.
• API access and consulting — programmatic access to semiconductor data, and direct engagement with the research team.

Choosing Between Tracks

• If you need to follow industry developments and occasional analysis, start with the newsletter.
• If you need SKU-level forecasts, installed base, or datacenter capacity data for planning or investment work, the institutional models are the relevant product—plan for a sales conversation.
• If you need to pipe data into your own systems, ask about API access during that conversation.

Common Pitfalls When Using This Research

• Don't read bottoms-up forecasts as precise point estimates. They are conditional on equipment, wafer start, and packaging assumptions; when a constraint shifts, the forecast shifts. Check which input is binding before citing a number.
• Match model granularity to your question. The datacenter model answers power and capacity questions; the accelerator model answers shipment questions. Using one for the other's purpose will mislead.
• Verify access terms before budgeting. The available site information confirms a subscribe link and an institutional login but does not state prices or whether any content is free.

For a quick orientation, the site's own framing is the clearest summary: the definitive source on the world's most important industry, built on models that start from production and end at deployment.

SemiAnalysis is an independent research outfit focused on the AI and semiconductor supply chain. According to its site, it publishes a newsletter read by 318,000+ subscribers and offers institutional products: fourteen models, open benchmarks, and advisory services. It's relevant if you need bottoms-up forecasts of AI accelerator and HBM demand, datacenter capacity and power tracking, or inference economics — and less relevant if you want general tech news or consumer hardware reviews.

What SemiAnalysis actually publishes

The site organizes its work into a few distinct tracks:

The newsletter is the entry point; the models and advisory are the institutional layer. The site separates these clearly, so you can start with the free-to-read newsletter and only move to the models if your work needs the underlying data.

The models, in more detail

The page evidence describes three model families with specific coverage. These details matter because they tell you whether the research matches your question.

Accelerator & HBM Model

• Bottoms-up demand forecast for AI accelerators and high-bandwidth memory across Nvidia, AMD, Google, AWS, Meta, Microsoft, Intel, and every major Chinese hyperscaler.
• Customer-level shipments and installed base for 60+ accounts.
• HBM tracked down to vendor, generation, layer count, and price.
• Because shipments are built up from equipment requirements, wafer starts, and packaging orders, supply-chain revenue can be estimated directly from production.
• Forecasts to 2030, history back to 2021.

Datacenter Industry Model

• Global capacity tracker covering 6,500+ structures: hyperscaler buildouts, colocation markets, power constraints, and regional deployment trends through 2030.
• Per-structure power draw, lease structures, and planned versus operational capacity.
• More than 200 companies covered, with the end customer behind each megawatt.
• Demand side reconciles accelerator shipments and install base by operator against power availability, showing which operators and regions run short of power, and when.
• Delivery tracking compares promised versus delivered MW quarterly.

Tokenomics Model

• Unit economics engine for AI serving: bottoms-up token throughput, ROI on AI spend, and translation of revenue into future accelerator demand by hyperscaler.

How to decide whether it fits your work

Use these conditions as a quick filter:

• Choose SemiAnalysis if you need SKU-level accelerator or HBM forecasts, structure-level datacenter power modeling, or inference economics — and you want the methodology visible through open benchmarks and tooling.
• Consider the newsletter alone if you want regular analysis without committing to institutional data products.
• Look elsewhere if your question is about semiconductor manufacturing processes in the abstract, consumer devices, or general AI policy; the coverage here is explicitly supply-chain and demand-modeling oriented.

One practical note: the site lists a newsletter subscription link and an institutional login, but the provided material does not state prices or whether any content is free versus gated. Check the subscribe page directly before assuming access terms.

What to verify before relying on it

The models are forecasts, and the site itself frames coverage as projections to 2030 with history to 2021. If you plan to cite numbers in a decision, confirm the model version and vintage you're using, and treat the bottoms-up build (equipment → wafer starts → packaging → shipments) as the thing to audit — that chain is where forecast error compounds. The open tooling and benchmarks exist precisely so you can inspect assumptions rather than accept headline figures.