Teradyne, Inc. has announced a production-ready system that delivers enhanced burn-in capabilities on its Titan HP platform. Traditional burn-in holds devices in banks of dedicated ovens at a controlled air temperature for hours before they move on to functional testing.
Teradyne Titan HP instead holds each device at its target junction temperature while running a realistic device workload, so reliability stress functions are optimised and thermal controls are enabled at the per-device level. Titan HP can be used for system-level testing or reliability testing, optimising capital investment. For reliability testing, every device carries its full stress-and-test history in a single record.
“Our customers are being asked to ship high-power AI silicon at reliability levels that used to belong to automotive,” said Jason Zee, vice president and general manager of the Integrated Systems Test division at Teradyne. “Moving reliability stress testing onto a system-level test platform means optimised reliability, which can lower testing costs, improve time to market and deliver better data on every device shipped.”
Per-site thermal control, tuned to each device’s thermal and force zones and paired with high per-device kilowatt power delivery, lets Titan HP sustain reliability stress conditions on parts whose power draw and heat dissipation strain a conventional oven process. The platform’s asynchronous slot architecture loads and tests each slot independently, so a long stress run on one device does not leave the rest of the system idle, and individual sites can be serviced while the system continues to operate.
Test programs run on the Teradyne Atlas software platform, which lets engineers import existing programs or build new ones and port them to other Teradyne system-level test systems. Teradyne’s per-device power roadmap extends to the levels future AI accelerators are expected to require, so the same platform carries customers forward as device power grows.
Burn-in has long been standard for automotive and other high-reliability silicon. AI accelerators and cloud infrastructure devices are now entering the same category, driven by the high cost of a single failed accelerator in a deployed rack. That demand is arriving as capital, floor space and handling capacity for dedicated oven banks come under pressure in high-volume manufacturing and OSAT facilities.
Enhanced burn-in extends Teradyne’s test portfolio, which spans the AI device supply chain from wafer to data centre. As AI silicon moves through wafer probe, package testing, system-level testing and now reliability stress testing, Teradyne covers every device at every insertion.
Read other recent instrumentation news: https://instrumentation.co.uk/category/news/
