Method
How design-in signals are detected.
A signal is a piece of public evidence that bears on one of two questions: could this part serve this product, and is the design open. Different kinds of page answer different questions, and no single kind answers both.
This page lists the sources in use, what each one can establish, and what a test run returned.
What each source can establish, and what it cannot.
Status as of September 2026. Tested means run on real products with the results checked by a person.
| Source | Can establish | Cannot establish on its own | Status |
|---|---|---|---|
| Datasheets and product pages | External interfaces, power, temperature range, throughput | What is inside the product | Tested |
| Regulatory filings with internal photos | A specific device on a specific board, at filing date | Whether the current revision still uses it | Tested |
| Teardowns and chip identifications | The device inside a shipping product | Industrial products, which are rarely torn down in public | Tested |
| Job postings | The device families and tools a team works with; that a team is hiring | Which product a device is in; that a purchase is planned | Tested |
| Trade press and launch articles | That a product exists, sometimes the platform it is built on | That component selection is open | Tested, low yield |
| Code repositories | Board support, drivers, reference designs in public use | Commercial products, which are mostly private | Not yet tested |
| Forums and communities | Problems engineers have with an incumbent part | Anything about a named product without corroboration | Not yet tested |
Alongside public sources, the manufacturer's own knowledge of where competing parts are used is the most direct evidence there is. It fills the same field for products that no public page has opened.
The kind of page matters more than the number of searches.
In one day of structured searching, queries for a part number returned pages selling the part. Queries restricted to the places where someone opens a product and names the chip returned products containing it.
Of roughly 6,400 search results collected, about 520 pages named both a specific product and the programmable device inside it. About 150 of those were shipping products and not evaluation boards, and about 100 were in industrial application classes. None of these has been hand-checked yet.
Targeted searching found the incumbent device for two and a redesign signal for one. Three stayed unknown. No teardown, no filing, no public answer.
Public search can fill the missing field for some products and cannot for others. The rate across a full catalogue is not yet known, and it decides whether the result is a decision or a filter.
How a signal is allowed to be used.
These hold for every record.
- Every value carries the text it came from and the page it was found on.
- Evidence about a company is kept apart from evidence about a product.
- A value that is not stated is recorded as unknown, never filled from general knowledge.
- Technical fit and timing are assessed separately.
- An unknown is not a mismatch.
One page per source.
- What the source is and where it is found
- What it has established in testing
- Where it misleads
Datasheet signals · FCC filing signals · Job posting signals · Teardown signals · Design-in intelligence · FPGA product-fit screening
Ask what the evidence can establish for your part