A Novel Massively Parallel Testing Method Using Multi-Root for High Reliability
| Autor: | Haksong Kim, Yong Lee, Sungho Kang |
|---|---|
| Rok vydání: | 2015 |
| Předmět: |
Very-large-scale integration
Engineering business.industry Test compression Hardware_PERFORMANCEANDRELIABILITY Automatic test equipment Parallel processing (DSP implementation) Embedded system Hardware_INTEGRATEDCIRCUITS Wafer testing System on a chip Electrical and Electronic Engineering Safety Risk Reliability and Quality business Massively parallel Test data |
| Zdroj: | IEEE Transactions on Reliability. 64:486-496 |
| ISSN: | 1558-1721 0018-9529 |
| Popis: | Wafer testing (wafer sort) is used in the semiconductor industry to reduce test costs. High parallelism is important to reduce test application time. However, increasing parallelism is becoming more difficult because the elements that drive test costs are increases in the pin count of the system on chip (SOC), and required automated test equipment (ATE) channels. While the need for parallelism has been growing, a reliability problem in which fault distribution causes good devices under test (DUT) to be improperly tested is becoming a concern. To achieve high parallelism and reliability, we propose a novel, massively parallel testing method using multi-root. In addition, we develop a test setup for setting the root-DUT location and the path of all DUTs. Using the proposed wafer testing method, test data can be transferred between the ATE and dies using multi-roots. The experimental results using ITC 02 SOC benchmarks show that the proposed method can reduce test costs up to 90%, and achieve nearly 94.84% test reliability without affecting yield. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|