CAST Work Group updates from SEMICON West 2009
To View agenda, please click here
Test Cell Communications Working Group
Meeting Date: July 15 2009
Karl Stuber - Semi
Somnath Sen -Consultant
Mark Roos - Roos Instruments
Anant Anavkar - iSoft Group
Joel Taylor - Verigy
Lonny Plummer - Kinesys Software
Emilio Salvioni - Sallan Engineering
John Bearden -Consultant
Kevin Fetterly - Apria Tech
The meeting started with a review of the April 1 2009 meeting at the Semi facility. This was followed by a statement of need for the outcome of the group. Initially, it was felt that a common method of communicating to probers and handlers was needed. However, the general consensus was that we would not influence the large installed base of this type of equipment. The proposal put on the table by Mark Roos was to have a forward-looking view of the over all test cell and define how we can influence or drive a data-centric view of the test cell as opposed to an equipment-centric view. This put a new light on the direction of the group and the remainder of the meeting was focused on "fleshing out" what a data-centric test cell would look like.
Date-centric Test Cell
The discussion began with the concept of a logical level model of the test cell.
- the main logical functions in the test cell
- the question "where is it done today and is this the best place for it to be done?". The following was a first attempt to define these logical blocks:
Datalogging - This point was broken into two possibilities
- "Critical" Data - required for binning - this type could be processed immediately
- "Interesting" Data - not required for binning - process latest for things like statistical analysis
The goal is to off-load the test cell from a great deal of datalog processing that simply is not needed at that point in time. This would increase the efficiency of the test cell while maintain (or possibly increasing) the amount and quality of data supplied by the test process.
- Central control (i.e Cell Controller) versus decentralized functions (i.e. tester-centric model).
- The goal of this new model would be to define expanded capability and added value that would drive customers and suppliers alike to implement. Once established, legacy equipment could be accommodated in the new model via software modules or adaptors of some kind.
In order to begin putting some structure around this idea, we need to put together a typical device flow as it goes through the cell needs to be derived. Not only do the physical steps need to be understood but we also need to understand what data is collected at each step, which part of the test cell collects it and where does it end up. This view of the test cell will drive a state-machine-like view of the test cell from which a data-centric model can be derived.
Possible benefit with other working groups
It was identified that the data-centric view of a test cell may have some over lap with the Yield Enhancement WG. It may be possible that output from the Yield enhancement tools may be used as input to the data-centric test cell to enable yield-based or yield-influenced test flow. The other benefit of interfacing with the Yield Enhancement WG is assistance in defining the granularity of the data tasks that need to be defined in the data model - how "small" does a task need to be to be useful to the YE WG? This was not discussed further in detail, only that it may be worthwhile to see if there is mutual benefit in the data view of a test cell.
The main action that came out of the meeting was for the attendees to derive their view of the flow of a device through an existing test cell and attempt to define the "what, where and when" of the data steps in the cell (as it is done today).
The group concluded that monthly conference calls would be the appropriate frequency starting in the September 2009 time frame. Face to face meetings would be arranged, where appropriate, to coincide with major industry events (SEMI conferences, ITC etc.). Plans are beginning to come together for a face-to-face meeting of this WG on October 6 2009 at the SEMI facility in San Jose
STDF Fail Datalog Working Group
Where: Moscone Center, San Francisco
Bill Price – T2C2
Wes Smith – Galaxy Semi
Cy Hay – Synopsys
Keith Arnold – Pintail
John Beardon – Beardon Consulting
Thierry De Villeneue – ST Micro Electronics
Ajay Khoche – Independent
- Introductions and Welcome
- Scan Fail datalog Status update
– Deployment status
– Validation Tool usage
- Memory Fail datalog Status update
Migration of Efforts under CAST/SEMI
-Ownership of STDF
- Scan fail datalog update
- The standard is ready since Nov 2008
Change in Website: The new website for the group is now
- Please note that this would change again as we move the effort under SEMI/CAST
- Currently supported by verigy
- Teradyne plan to support starting 8.0
- Advantest need customer pull
- EDA companies are also waiting for customer pull
- Deployment efforts have been affected by the downturn
STDF Navigator tool for validating readers and writers of STDF 2007 is ready
- Many companies are currently using the version released in Feb 2009
- Currently discussions are underway with Verigy to regain rights to distribute and support.
- A notice will be sent out about the availability and direction to get access once the discussions are settled.
- The tool will be available free of charge.
- Memory Fail datalog Update
- The technical discussions are complete
- The group is ready for ballot
- Migration of Efforts under SEMI/CAST
The Memory WG has agreed to conduct ballot under SEMI/CAST for following reason
- Broader reach
- World wide membership
- About 400 active members
- Comprehensive review
- Better chances of adoption
- Group needs a sponsor anyway
Impact of Moving under CAST/SEMI
- The efforts needs to be approved by the CAST committee
- We need to follow CAST/SEMI process for ballot and approval
- Need to follow SEMI Timelines
- A formal request for the same has been made to North American ATE committee under SEMI/CAST
- Work (e.g. ballot, group formation etc) needs to be approved by the committee
- Next meeting is at SEMI West
- 2 wks wait after the approval
- Ballot (30 day)
- Adjudicate (By NA-ATE committee)
- Procedural review
- Publication if approved
The details of the technical discussions are in the attached presentation.
Docking & Interface Working Group:
Date: Wednesday, July 15
Location: Marriott Hotel, San Francisco
Doug Lefever Advantest
Edward Smith Consultant
Kosuke Miyao Verigy
Chris Beronio W.L. Gore
David Oka LOA Technology
Kevin Fetterly Apria Technology
Keith Imai Consultant
Rene Monnet STMicroelectronics
Welcome and Introductions
SEMI Antitrust Guidelines
History of DIWG
The DIWG does not have a chair. Paul Roddy led the meeting. Paul started the meeting with introductions and mention of the SEMI Anti Trust Rules. Paul then reviewed over the history of the DIWG (please see the 7/15/09 DIWG meeting presentation for details). Paul also reviewed over the notes from the 4/1/09 DIWG meeting. Paul then talked about future plans for the DIWG. Paul asked for a volunteer to chair this WG. Paul then opened the floor for discussion. It was agreed that the semi equipment “users” should be the ones to drive this WG effort. Everyone agreed that we needed additional members for this WG, to achieve critical mass. Ed Smith took an action item to help us recruit additional members. Paul took an action item to contact all current CAST members and ask them to identify a representative from their company to serve on this WG. We will identify and recruit members needed to achieve critical mass, before starting regular conference calls. The next face meeting is tentatively schedule to be held during ITC in Austin, Texas, during the first week of November.
PTIM Working Group
Chris White (NI)
Scott Savage NI
Tod Gutfeldt Advantest
Todd Luppin Advantest
Rudra Kar Advantest
Kegin Fetterly Apria Tech
David Oka LOA Tech
Thijs Waardenburg Aeroflex
Edward Smith Salvador and Smith
Bruce Parnas Event Based Technologies
Anil Godbole Event Based Technologies
Glen Gomes Event Based Technologies
- Overview of PTIM history and charter (30 min)
Technical overview of two industry-standard instrumentation specifications
- PXI and LXI (1 hour)
- General discussion (30 min)
About half the room was new to PTIM. Went over last few years of PTIM history to increase awareness of problem statement, types of PTIM devices, and charter. There were general questions but no big issues raised.
We then dove into a technical overview of PXI and LXI. This was an action item from the last PTIM meeting, and was an important step for consensus back when PTIM was part of STC. The purpose was to increase awareness of the form-factor, general capabilities, and what is and is not defined in the standards' specifications. Through this discussion, it can become more
apparent what the pros and cons are of these industry standards as it relates to ATE. There were plenty of clarification questions, but no big issues raised.
Since no difficult issues were brought up during the first hour and a half, we looked at the progress to date on the guidelines document. About half of the document is drafted, but several key sections remain to be developed. Much of these remaining areas will require strong participation from some "regular" PTIM members. we discussed ideas on how best to move
forward, and it was agreed that we need more regular meetings outside of the "official" working group meetings where we can make progress.
I raised the point that one of the challenges with the PTIM group was its lack of continuity in its member base. To date, NI and Aeroflex (both PXI vendors) have been working on the spec. However, since some 2 of the 3 PTIM instantiations would require development from the ATE companies, it seems important that there be ATE participation in the creation of the guidelines document. Most people around the room agreed that they would like to be involved, but we left the meeting unclear as to the next steps.
I spoke with Paul Roddy after the meeting and asked if we (Luke and I) could meet with him after SEMICON West to discuss how best to move forward.meeting scheduled for September.