HS 84, 85, 90
Customs records across the 3 HS chapters it hardware trades under — computers and servers and the rest of the category, with named buyers, suppliers, quantities and prices.
IT Hardware & Computers covers computers and servers, storage, networking equipment, peripherals and displays, data centre infrastructure, components and sub-assemblies. All of it crosses borders under chapters 84, 85, 90 of the Harmonized System, which is the only definition of the industry that customs — and therefore the data — actually recognises.
| # | Market | Value | Year |
|---|---|---|---|
| 01 | United States | $1.14T | 2024 |
| 02 | China | $887.9B | 2024 |
| 03 | Hong Kong | $471.1B | 2024 |
| 04 | Germany | $417.6B | 2024 |
| 05 | Mexico | $259.4B | 2024 |
| 06 | Singapore | $240.7B | 2024 |
| 07 | South Korea | $212.6B | 2024 |
| 08 | Japan | $206.2B | 2024 |
| 09 | Taiwan | $205.6B | 2024 |
| 10 | United Kingdom | $193.8B | 2024 |
| 11 | Netherlands | $182.7B | 2024 |
| 12 | France | $181.0B | 2024 |
| # | Market | Value | Year |
|---|---|---|---|
| 01 | China | $1.57T | 2024 |
| 02 | United States | $572.6B | 2024 |
| 03 | Germany | $521.9B | 2024 |
| 04 | Hong Kong | $476.2B | 2024 |
| 05 | Taiwan | $358.6B | 2024 |
| 06 | South Korea | $310.3B | 2024 |
| 07 | Singapore | $292.5B | 2024 |
| 08 | Japan | $262.9B | 2024 |
| 09 | Mexico | $243.6B | 2024 |
| 10 | Netherlands | $206.4B | 2024 |
| 11 | Italy | $178.5B | 2024 |
| 12 | Vietnam | $169.3B | 2023 |
Source: UN Comtrade, official statistics reported by national authorities. Figures are country-level aggregates, not shipment records.
Open a chapter for its own market table, partner countries and multi-year values. The chapter pages are where a product-level search starts.
What the record shows
Two sides of the same declaration, and they answer different questions.
Import records are your demand map: data centre and enterprise buyers import in bursts that track their build cycles.
Export records are your supply map: model-level detail is often in the description, which makes competitive mapping unusually precise.
What India reports across the it hardware chapters, on each chapter’s latest filed year. Use it to size the domestic base before you go looking for a counterparty in it.
| HS | Chapter | Exports | Year |
|---|---|---|---|
| 85 | Electrical machinery and equipment and parts thereof | $39.4B | 2024 |
| 84 | Nuclear reactors, boilers, machinery and mechanical… | $32.0B | 2024 |
| 90 | Optical, photographic, cinematographic, measuring… | $5.0B | 2024 |
| HS | Chapter | Imports | Year |
|---|---|---|---|
| 85 | Electrical machinery and equipment and parts thereof | $83.5B | 2024 |
| 84 | Nuclear reactors, boilers, machinery and mechanical… | $60.8B | 2024 |
| 90 | Optical, photographic, cinematographic, measuring… | $13.5B | 2024 |
There is no single tariff line called it hardware. Customs classifies goods, not industries, so the trade that a it hardware business actually does is spread across chapters 84, 85, 90. Search one of them and you measure a slice; read them together and you measure the category. This is not a subtlety — for a multi-chapter industry, a single-chapter view can understate the market by more than half, and every conclusion built on it inherits the error.
The practical consequence is that market sizing, competitor tracking and buyer discovery all have to be run across the full chapter set. A competitor that appears to have shrunk may simply have shifted its mix from one chapter to another. A market that looks saturated in one chapter may be wide open in the adjacent one. The chapter pages linked above are the detail; this page is the view that holds them together.
| Chapter | Covers | Read it for |
|---|---|---|
| 84 | Nuclear reactors, boilers, machinery and mechanical applia | Product-level markets, partners and multi-year values |
| 85 | Electrical machinery and equipment and parts thereof | Product-level markets, partners and multi-year values |
| 90 | Optical, photographic, cinematographic, measuring, medical | Product-level markets, partners and multi-year values |
Every importer in the record is already buying the category. That is the whole advantage over a directory, and it is worth being precise about how to exploit it. Start by fixing the chapters, then narrow to markets you can actually service — certification, logistics and price all have to work before volume means anything. What remains is a list of companies with a demonstrated, dated, quantified need.
Rank that list on behaviour rather than on size. A buyer clearing consignments in most months has an operating requirement; a buyer with one large consignment two years ago has a project that has finished. A buyer already using several suppliers is more approachable than one that has bought from a single origin for a decade. In it hardware specifically, data centre and enterprise buyers import in bursts that track their build cycles.
Read the export flow of the same chapters and the question inverts. A genuine manufacturer ships consistently, across several destinations, over years, at consignment sizes that are stable enough to imply a production process behind them. A merchant exporter ships across unrelated chapters without a matching input trail. Neither is disqualifying, but you should know which you are talking to before you negotiate terms, because their cost structures are not comparable.
One point specific to this category: model-level detail is often in the description, which makes competitive mapping unusually precise. That is the kind of detail that separates a useful shortlist from a long one, and it is not visible in any self-reported profile.
Declared value divided by declared quantity gives a unit value, and aggregated carefully it becomes a price band you can hold a counterparty to. Carefully is the operative word: compare at the tightest published digit level, within one period, from one origin, in one unit of measure, with the outliers removed. A comparison that fails any of those controls will be dismissed the moment you put it on a table, and rightly.
Read the spread rather than the average. A tight band means the price is structural and your negotiating effort belongs elsewhere — payment terms, lead time, minimum order quantity. A wide band inside a single HS line usually means a grade difference that the declaration does not capture, and asking a supplier to name that difference is a more productive conversation than asking them to match a number.
Almost every traded category has a buying calendar, and reading a single period against no baseline produces confident, wrong conclusions. Compare like periods year on year, use at least three years before calling anything a trend, and drop the most recent one or two periods entirely — they are still filling in. A buyer who appears to have stopped buying has very often just been read too early.
These are official aggregates reported to UN Comtrade: country-level, annual, and reconciled across reporters. They establish the shape of the market. The shipment-level records that name a specific importer or exporter are a different dataset, filed with national customs authorities, and they are what a buyer or supplier list is actually built from.
Every importer of your it hardware lines in a chosen market, ranked by shipment count and declared value.
Confirm shipping history, destination spread and consistency before a first order rather than after it.
Compare declared values by origin so a quotation is argued from the record instead of from a hunch.
Follow a named company across the chapters — what it moves, where from, and how the volume is trending.
Measure real it hardware import demand in a country before committing to registration or certification.
Work back from real shipment descriptions to the classification customs actually accepted for goods like yours.
Related
It is the customs record for every consignment that clears under the HS chapters it hardware files against — 84, 85, 90. Each row names the importer and exporter, the product as declared, the quantity, the value and the ports used.
IT Hardware & Computers spans chapters 84, 85, 90. No single chapter contains the industry, which is why searching one code at a time under-counts the market — the chapters have to be read together.
Filter import records to the chapters above and to the markets you can service. The companies that recur period after period are the accounts worth approaching; data centre and enterprise buyers import in bursts that track their build cycles.
Read the export side of the same chapters. A genuine manufacturer ships consistently to several destinations over years. Note that model-level detail is often in the description, which makes competitive mapping unusually precise.
Yes. Every market with a country page carries its own chapter breakdown, and the shipment-level records are filterable by country, port and date range.
Because customs classifies goods, not industries. IT Hardware & Computers is spread across 3 chapters, so a single-chapter search measures a slice of the category and inherits that error into every conclusion built on it. Read the chapters together.
Divide declared value by declared quantity, then control the comparison: same tariff line, same period, same origin, same unit of measure, outliers removed. Read the spread rather than the average — a wide spread inside one line is usually a grade difference the declaration does not capture.
Most traded categories have a buying calendar, so compare like periods year on year rather than month on month, use at least three years before calling anything a trend, and exclude the most recent periods, which are still filling in.
The tables on this page are official country-level aggregates reported to UN Comtrade — they establish the shape of the market. Shipment-level records are filed with national customs authorities and name individual importers and exporters; that is the dataset a buyer or supplier list is actually built from.
Tell us which of the it hardware chapters you trade and we will send live customs records for it, drawn from the market you name.