Silent Data Corruption by 10x Test Escapes Threatens Reliable Computing
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908497256382464 |
|---|---|
| author | Mitra, Subhasish Banerjee, Subho Dixon, Martin Govindaraju, Rama Hochschild, Peter Liu, Eric X. Parthasarathy, Bharath Ranganathan, Parthasarathy |
| author_facet | Mitra, Subhasish Banerjee, Subho Dixon, Martin Govindaraju, Rama Hochschild, Peter Liu, Eric X. Parthasarathy, Bharath Ranganathan, Parthasarathy |
| contents | Too many defective compute chips are escaping existing manufacturing tests -- at least an order of magnitude more than industrial targets across all compute chip types in data centers. Silent data corruptions (SDCs) caused by test escapes, when left unaddressed, pose a major threat to reliable computing. We present a three-pronged approach outlining future directions for overcoming test escapes: (a) Quick diagnosis of defective chips directly from system-level incorrect behaviors. Such diagnosis is critical for gaining insights into why so many defective chips escape existing manufacturing testing. (b) In-field detection of defective chips. (c) New test experiments to understand the effectiveness of new techniques for detecting defective chips. These experiments must overcome the drawbacks and pitfalls of previous industrial test experiments and case studies. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_01786 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Silent Data Corruption by 10x Test Escapes Threatens Reliable Computing Mitra, Subhasish Banerjee, Subho Dixon, Martin Govindaraju, Rama Hochschild, Peter Liu, Eric X. Parthasarathy, Bharath Ranganathan, Parthasarathy Hardware Architecture B.7; C.0; C.5 Too many defective compute chips are escaping existing manufacturing tests -- at least an order of magnitude more than industrial targets across all compute chip types in data centers. Silent data corruptions (SDCs) caused by test escapes, when left unaddressed, pose a major threat to reliable computing. We present a three-pronged approach outlining future directions for overcoming test escapes: (a) Quick diagnosis of defective chips directly from system-level incorrect behaviors. Such diagnosis is critical for gaining insights into why so many defective chips escape existing manufacturing testing. (b) In-field detection of defective chips. (c) New test experiments to understand the effectiveness of new techniques for detecting defective chips. These experiments must overcome the drawbacks and pitfalls of previous industrial test experiments and case studies. |
| title | Silent Data Corruption by 10x Test Escapes Threatens Reliable Computing |
| topic | Hardware Architecture B.7; C.0; C.5 |
| url | https://arxiv.org/abs/2508.01786 |