what is burn-in testing
Burn-in testing is a reliability screening step where chips are operated under elevated temperature and voltage stress for an extended period before shipment, deliberately accelerating early-life failures so weak parts fail in the factory rather than in a customer's product.
```flowchart
{
"rows": [
{ "type": "nodes", "items": [
{ "title": "Chip passes standard functional test", "sub": "but early-life weak parts can still slip through", "tone": "neutral" }
]},
{ "type": "arrow" },
{ "type": "group", "title": "Burn-in: stressed operation over time", "items": [
{ "title": "Elevated temperature and voltage accelerate aging", "sub": "weak parts fail now instead of in the field", "tone": "orange" }
]},
{ "type": "arrow" },
{ "type": "nodes", "items": [
{ "title": "Survivors ship with much higher reliability confidence", "sub": "early-failure risk has already been screened out", "tone": "green" }
]}
]
}
```
**Burn-in testing exists because a chip's failure rate isn't constant over its lifetime — it follows a well-known pattern where early-life failures are disproportionately common, and burn-in deliberately forces those failures to happen before shipment.** Chip reliability commonly follows a "bathtub curve," with a higher failure rate early in life, a long stable middle period, and rising failures again near end of life; burn-in testing exploits the early-life portion of that curve by running chips under elevated temperature and voltage stress long enough to force out weak parts that would otherwise have failed early in a customer's actual product.
```svg
```
```svg
```
| Aspect | Standard functional test | Burn-in testing |
|---|---|---|
| Duration | Brief, checks correctness | Extended, checks durability under stress |
| Conditions | Normal operating conditions | Elevated temperature and voltage |
| Purpose | Confirms the chip works | Forces out early-life failures before shipment |
| Typical use | Every chip, always | High-reliability applications (automotive, aerospace, server) |
**Burn-in testing is used selectively rather than universally, since the added time, equipment, and energy cost only make sense for applications where early-life failure is especially unacceptable.** Consumer chips with lower reliability requirements and tighter cost pressure often skip extensive burn-in, while chips destined for automotive, aerospace, medical, or server applications, where an early-life failure could be costly or dangerous, are much more likely to undergo dedicated burn-in screening as a standard part of qualification.
**Determining the right burn-in duration and stress conditions requires balancing thorough failure screening against the real cost of extended, elevated-stress testing time.** Too little burn-in time or stress risks letting weak parts slip through undetected, while excessive burn-in adds meaningful cost and can itself wear down otherwise healthy parts — semiconductor reliability engineers calibrate burn-in conditions carefully based on a chip's specific failure-rate data and reliability requirements.
**Burn-in testing is one part of a broader chip reliability qualification process that also includes other accelerated stress tests targeting different failure mechanisms.** Beyond burn-in's focus on early-life electrical failures, a full reliability qualification program typically also includes tests targeting mechanisms like electromigration, thermal cycling, and moisture sensitivity — burn-in specifically addresses the early-life portion of a chip's overall reliability profile rather than every possible failure mode.
Read burn-in testing through a stress-test-now-not-later lens: by deliberately running chips hard under elevated stress before they ever reach a customer, burn-in forces the failures that would otherwise show up early in the field to happen instead in a controlled factory setting, where a weak part can simply be discarded rather than causing a real-world failure.