Summer learning series seminar: Improving the throughput of volume scan diagnosis
Performing volume scan diagnosis on today’s large, advanced-node designs puts outsized demands on turn-around-time and compute resources. Mentor offers a technology called dynamic partitioning that will maximize diagnosis throughput while performing ever more demanding volume scan diagnosis.
The live, one-hour, online seminar Tessent DDYA – Improving the throughput of volume scan diagnosis by 10X using dynamic partitioning is presented by Mentor expert Jayant D’Souza.
When:
- July 14, 2020, at 5:00 PM US/Pacific.
- July 15, 2020, at 8:00 AM US/Pacific.
If you can’t make either of the live sessions, register anyway and you’ll get the link to the recorded session afterward.
Where: Live, online presentation, and Q&A with the expert.
Cost: An hour of your time.
Traditional volume scan diagnosis takes full design netlists as the input and requires many parallel simulation runs, but each run requires a similar memory resource. For a large design, you would need several large memory machines to perform parallel runs. You might have 20 machines, but only one is big enough to handle the job and the rest sit idle.
The design size can be reduced dynamically. For a failing die, there are some patterns that fail, others that pass. Failing patterns activate the defect location, so only the part of the design containing the failing pattern might be needed.
Fail logs from the tester are analyzed by the Tessent Diagnosis Server, which works on your compute grid to both analyze and distribute the diagnosis processes. The server partitions the fail log files into smaller files containing just the information needed to find the defect, with a size cap of 20% of the full design size. The Tessent Diagnosis Server also includes analyzers that perform the actual diagnosis on the partitions and a monitor that coordinates between the partitioners and analyzers and intelligently distributes the diagnosis processes.
Using the multiple smaller input files with just the relevant fail log data means that more processes (an unlimited number, actually) can be run in parallel, using a wider range of CPUs, and getting results faster and more efficiently than ever.
Dynamic partitioning addresses hardware resource limitations by greatly reducing the memory footprint of the diagnosis process performed by the analyzers. Typically, the dynamic partitioning leads to a 5X reduction in memory and a runtime reduction of 50% per diagnosis report.
The seminar covers:
- The basics of scan test and scan diagnosis
- How scan diagnosis is being used for yield analysis and to guide failure analysis
- How volume scan diagnosis data is being used to construct defect paretos
- How new technology can make efficient use of hardware resources to enable faster availability of this data
