Standard for fan-out panel size ready to ballot

By Richard Allen

The arrival of Fan-Out Panel Level Packaging (FO-PLP) appears to be at a perfect time: This technology will leverage processes developed for Three Dimensional Stacked Integrated Circuits (3DS-IC) as well as panel processing technologies developed for industries such as solar panels and large-screen TVs.  In this combination, FO-PLP promised the improved performance of 3DS-IC, without the expense. There was just one problem…

That problem is the size of the panels to be processed. As different companies developed FO-PLP processes, they chose panels sized to meet certain technical or business goals, or chose a size based on familiarity. So, processes were being developed for more than ten sizes, each of which had one or more companies championing them.

For people in the wider semiconductor industry, the development of many processes, each with a unique panel size brought a feeling of déjà vu, reminding them of the 1970s, when each device manufacturer created their own specification for wafer size, forcing them to manufacture their own wafer processing equipment since no external manufacturer was willing to produce tools usable only by a single customer.

SEMI responded by developing an industry consensus silicon wafer standard – which described basic parameters, including diameter and thickness – to resolve the issue. Almost overnight the landscape changed, and new tool manufacturers sprung up, enabling the incredible growth that has persisted over more than 40 years.

Recently, Cristina Chu (TEL NEXX) presented the state of FO-PLP to the North America Chapter of the SEMI Three-Dimensional Packaging and Integration (3DP&I) Technical Committee, suggesting that the Committee develop a single standard dimension that would enable the technology to move into high-volume manufacturing.

The Committee began by surveying the industry to determine the interest level in such a standard as well as its contents.  A key finding came in response to the question “Would you support a standardized panel size?” Overwhelmingly, over 70 percent of the respondents supporting the idea for the standard, with less than 2 percent opposed. The survey also asked if other parameters should be standardized and, if so, which parameters. Majority responses pointed to edge profile, flatness, and warp, prompting the 3DP&I Committee to immediately form the FO-PLP Panel Task Force (TF) to develop such a standard. Chu and Richard Allen (NIST) agreed to chair the TF and respondents to the survey were asked to participate as TF members.

The TF initially decided to follow the model of SEMI M1, Specification for Polished Single Crystal Silicon Wafers, and write the document as a purchase specification. The purchase specification would indicate a limited number of mandatory parameters, identified as those that serve as bottlenecks to the development of a FO-PLP ecosystem. Parameters that were not perceived as bottlenecks but might be useful for implementing a FO-PLP process would be included as optional.

Working under the SEMI Standards umbrella allowed the TF to take advantage of work done in the development of other standards, without having to recreate it from scratch. In particular, Flatness and Shape were repurposed from SEMI M1, ensuring consistent definitions of these parameters.

The TF could not come to consensus on how the other parameters should be categorized, so the decision was made to move the ordering table to a new Appendix as optional.

The TF will be balloting its first specification for panel substrate in the upcoming cycle, which opens September 5, 2018 (Cycle 7). The voting is open to all industry experts. Based on the feedback, the task force will continue to refine and otherwise improve the specification by incorporating other parameters that are critical to making FO-PLP a reality.

SEMI Standards development activities take place throughout the year in all major manufacturing regions. To get involved, join the SEMI International Standards Program at:

For more information regarding FO-PLP Panel Task Force activities, please contact Laura Nguyen at [email protected].

Richard Allen is a physicist in the Nanoscale Metrology Group in the Engineering Physics Division of the Physical Measurement Laboratory (PML) at the National Institute of Standards and Technology (NIST). 

Originally published on the SEMI blog.


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.

One thought on “Standard for fan-out panel size ready to ballot

  1. Dr. Dev Gupta

    The FO WLP format using reconstituted 300 mm wafers is cost wise competitive to large ( 40 mm sq. ) 2.5 d Silicon Interposers so long as the fine L/S/ Via Dias for such modules required can be delivered ( FO CoS by ASE ), But FO WLPs are uncompetitive to incumbent substrate based Flip Chip packages at package sizes over 15 mm sq. In order to reduce the cost of FO PLPs processing in large panel size is one option so that the area utilization for active devices can be increased even for large packages for small dies at adv. nodes w/ many I/Os.

    With the semiconductor industry still deeply uninterested in Tool development for 450 mm wafers, the Packaging community ( mainly Govt. funded educational Labs in Europe who IMHO instead of indulging in Process / Tool Integration should be doing fundamental research ! ) have jumped into the fray.

    Current Tool set ( litho, deposition ,.. ) infrastructure for Flat Panel Displays should be a good upper bound for FO PLPs. Safe to say though that at present PLPs would need to aim at much smaller panel sizes due to the warpage caused by mold compound which also affects die shift in the most common ” first Mold then RDL” flow. There are workarounds for both issues but they are expensive, thus defeating the goal for going from 300 mm reconstituted wafers to larger panels.

    Before jumping off to selecting mechanical details like panel size ( important as that dictates Tool design ) it would be better to first carefully characterize the dependence of key problems of panel processing ( e,g. warpage, effect on min. line widths,… ) as a function of various Mold Compounds for a range of Panel sizes, PUBLISH THE DATA at Conferences, then develop action plan for long lead time items like materials ( e.g. mold compounds )

Comments are closed.