AD-A277 686

SETA Support for the DARPA Microelectronics TechnologyInsertion Program of the Microelectronics Technology Office

U QUARTERLY TECHNICAL REPORT

I For the Period1 August 1992 to 31 October 1992

Sponsored by:Defense Advanced Research Projects Agency

Mlcroelectronlcs Technology OfficeMicroelectronics Technology Insertion Program

ARPA Order No. 8500 Program Code Nos. 2E20 and 2D10Issued by DARPA/CMO under Contract No. MDA972-92-C-0029

O T C -- Prepared By:

* ~ELF-CTE jTiAPR .- ,Daniel H. ButlerJr.

Approved By:

Ti dcmth." Robert Swistak ___

-Ms• document,,. has b:,,n approved• •forpubic elease and Sale; its= n is unlimitedi 15 February 1992 ___

* 1TIC QIJ~nU~ TGE

Booz-Allen & Hamilton Inc.

The views and conclusions contained In this document are those of the authorsand should not be Interpreted as representing the official policies, either expressed

or Implied, of the Defense Advanced Research Projects Agency or the U.S. Government.

Ioi l•

II,!

I

SETA Support for the DARPA Microelectronics Technology

Insertion Program of the Microelectronics Technology Office£QUARTERLY TECHNICAL REPORT

For the Period1 August 1992 to 31 October 1992

5 Sponsored by:Defense Advanced Research Projects Agency

Microelectronics Technology OfficeMicroelectronics Technology Insertion Program

ARPA Order No. 8500 Program Code Nos. 2E20 and 2D10Issued by DARPA/CMO under Contract No. MD' 372-92-C-0029

Prepared By: Accesion For

NTIS CRA&IDTIC TAB 7U. zarno c;d;ed r

Jus•tlf cation .......... ......5 Daniel H. Butler, Jr. ---B y ........................................ ........

Approved By: Dit.ibutvon/3 Avakability C.:d•.s

" - . Avail e2.d; orDist Special

Robert Swistak

3 15 February 1992

3 Booz.Allen & Hamilton Inc.

The views and conclusions contained In this document are those of the authorsI and should not be Interpreted as representing the official policies, either expressed

or Implied, of the Defense Advanced Research Projects Agency or the U.S. Government.I

£

3TABLE OF CONTENTS

PageNumber

1. INTRODUCTION ........................................................................................................ 1I2. TASK 3.1: Engineering and Technical Support Services ........................................ 1I3. TASK 3.2: Management and Administrative SupportS ..............S.............. 6I4. CON CLUSION S .............................O..........N....................................................................... 10l

I APPENDIX A: Trip Report, San Jose, California, 21-25 September 1992

I APPENDIX B:

Attachment 1: TriQuint Analysis, Nashua Meeting, 28 August 1992Attachment 2: Lockheed Sanders Analysis, Nashua Meeting, 28 August 1992Attachment 3: Lockheed Sanders Briefing Notes, Nashua Meeting, 28 August 1992

I APPENDIX C: Briefing On Digital GaAs Insertions: A Retrospective, 20 November 1992

I APPENDIX D- Schedule of Reviews and Demonstrations for the Transition of OpticalProcessors to Systems (TOPS) Program, October 1992 - December 1994

III

II

IREP-ORT DOCUMENTATIOPFPAGEr- JWX XNWo 00~4 go W sm Iw .uA T-ams Sowm., a#mem. - vnb~ l r sw 4111111 do soe3W~f MJ4M UW OUqUU mu141ý sJalil

15Fbur 19 ur.Teh e.- O A ug. :!92 - 31 Oct.92T&TLE AND SUBTITLE T- LPUNDING NUMIERSQuarterly Technical Report No.3 - SETA Support for theDARPA Microelectronics Technology Insertion Program of the MA7-2c02Microelectronics Technology office MA7-2C02

4AUThOFAS)Daniel H. Butler

-7. PtMORMWAG ORGANIZATION NAME(S) AND AOORISS(ESI L. PEPORMW ORASTIONBooz-Allen & Hamilton, Inc. REPOIRT NAUMBER4001 N. Fairfax DriveISuite 650N/Arlington, VA 22203N/

1 69. SpOISSOPNGJ/ MONITORIN AGENCY NAME(S) AND ADOESS(ES) I0. SPONSORING I MOINTORINGDARPA /MTO AGENCY REPORT -UII3701 N. Fairfax Drive3 Arlington, VA 22203-1714

11. SUPPLEMENTARY NOTES

K DISTRIBUTION STATEMENT A: Approved for public release;

disribtio isunlimited

3. Booz-Allen & Hamilton prvdsDARPA's Microelectronics Technology Yf-f-rc~e with

abroad range of SETA support under contract MDA972-92-C-0029. This report describeE

activities during the first quarter of this contract. The main programs supportedI were: the Digital Gallium Arsenide Insertion Program, the Transition of OpticalProcessors to Systems (TOPS), the Microelectronics Manufacturing Strategy (M"MST)Program, the Flexible, Intelligent Microelectronics Manufacturing Program (FIMM),andthe Artificial Neural Networks Technology Program (ANNT). In addition, support wasI begun to a new High Speed Circuit Design Program.

This report is organized by subtask areas in the statement of work, indicating

for each subtask the Task Objectives, General Methodology, Technical Results, andI Important Findings and Conclusions. The final section of this report presents asummary and conclusions, and the appendices present trip reports, meeting schedules,and some viewgraphs generated during this quarter of the contract.

14. SAUVT,A ITT, Transition of Optical Processors to Systems (TOPS), i.MmaO AEMicroelectronics, Artificial Neural Networks, MMST, FINN, ANN, 85R~ECD

Gallium Arsenide (GaAs)ILPCEODI-17. SECURITY CLASSIPIATION ILi SECURITY CLASSWCATMOE1. SECURTY CLOSSIFICATIO 20L LIMITATION OP ASRAMCTOP REPORT I OF ThtS PAGE I Of AASTRACT

Unclassified Unclassified Unclassified UL

I'4$1 ?540401-ZWSOSS Standard Form 296 (ROV. 2-89)Irmllg O &AS Souil. MO'S

US Scientific, Engineering, Technical Assistance Support

for the Microelectronics Technology Insertion Program of the MicroelectronicsTechnology Office

Defense Advanced Research Projects AgencyQuarterly Technical Report

For the Period1 August 1992 - 31 October 1992

I1. Introduction

I Booz.Allen & Hamilton Inc. provides the Microelectronics TechnologyOffice (MTO) of the Defense Advanced Research Projects Agency (DARPA) witha broad range of Scientific, Engineering, and Technical Assistance (SETA)support under contract MDA972-92-C-0029. This report describes activitiesduring the third quarter of the first year of this contract. The main programssupported were the Microelectronics Manufacturing programs, the Transition ofOptical Processors to Systems (TOPS) Program, the Digital Gallium ArsenideInsertion Program, and the Artificial Neural Network Technologies (ANNT)ft Program.

This report is organized by Statement of Work tasks. It describes the TaskObjectives, General Methodology, Technical Results, and Important Findings andConclusions for each subtask, and includes a final section of this reportconcerning contractual, administrative, and financial considerations.I2. Task 3.1: Engineering and Technical Support Services

This task involved three subtasks:

* Engineering and Technical Integration and CoordinationTechnical Studies and Assessments* Technical Documentation.

S2.1. Subtask 3.1.1: Engineering and Technical Integration and Coordination

Task Objectives. The objective of this subtask is to serve as liaison forvarious MTO programs, help identify and develop new programmaticopportunities, and explore pathways for technology transfer.

General Methodology. Booz.Allen serves as a primary liaison betweenMTO, funded contractors, and the technology community at large for theprograms supported under this contract. We maintain regular phone contactwith contractors; conduct on-site visits to contractor facilities; and participate in

2/15/93 MTO SETA Support: Quarterly Technical Report

I

Uconferences, workshops, and symposia. Booz-Allen monitors contractor

activities through informal on-site visits, workshops, and regularly scheduled In-Progress Reviews.

Technical Results, Findings, and Conclusions. During the third quarterof this contractual effort, Mr. Dan Butler, Booz.Allen program manager, andother Booz.Allen employees met with contractors and other parties involved inMicroelectronics Manufacturing and Digital Gallium Arsenide (GaAs) efforts tog serve as liaison for MTO and monitor progress toward each program's objectives.

Microelectronics Manufacturing activities included the following:

3 • Mr. Mark Connor attended the Microelectronics Processing '92Conference in San Jose, California on 21 September 1992 through 25 Septembera 1992. Appendix A contains a trip report from this conference.

* Mr. Graydon Larrabee reviewed the Sandia Contamination FreeManufacturing (CFM) Center.

* Mr. Larrabee made a presentation on Wet & Dry Cleaning toI Sandia National Laboratory.

• Mr. Larrabee attended the Sandia CFM - DARPA meeting.

I At the microelectronics processing conference, Mark Connor attended theAdvanced Techniques for Integrated Circuit Processing II Subconference onbehalf of Mr. Zachary Lemnios. Consideration of etching issues, especiallymagnetically enhanced reactive ion etching (MERIE), dominated nearly half ofthat subconference. The problem of microtrenching was also discussed andmerited further research, in Mr. Connor's opinion. Cost and technology issuesrelated to multi-chip module (MCM) manufacturing were also reviewed.

Graydon Larrabee reviewed the progress at the Contamination FreeManufacturing center at Sandia National Laboratory, and later made a technicalpresentation at Sandia on silicon-wafer cleaning. He also met withrepresentatives of Sandia, SEMATECH, and DARPA in Washington. Mr.Larrabee concluded that SEMATECH was driving the research center towardsshort term, incremental and improvement-type projects, a position in his opinionnot acceptable to DARPA or to forward looking American microelectronicsmanufacturers.

5 Digital Gallium Arsenide activities included the following:

On 28 August 1992, Mr. Dan Butler visited Lockheed Sanders toreview the current status of their digital GaAs Insertion project. Appendix Bcontains the trip report of this visit.

2/11M93 MTO SETA Support: Quarterly Technical Report 2

I

3 Mr. Butler attended the OH-58D Image Processor Program StatusReview in St. Louis, Missouri on 7 October 1992. The financial andprogrammatic status of this insertion effort were reviewed.

0 Mr. Kurt Hinds and Mr. Connor attended the 1992 IEEE GaAs ICSymposium at Miami Beach from 4 October to 7 October 1992. Mr. Hinds alsoattended a short course on Heterojunction Bipolar Transistor Integrated Circuit(HBT-IC) technology and applications during the conference.

In his position as contractor liaison between DARPA and LockheedSanders, Mr. Butler attended the Lockheed Sanders Status Review Briefing on the126B ECM GaAs Upgrade. Problems Lockheed Sanders was having with the firstset of TriQuint chips were discussed as well as proposed solutions; it was notedthat problems with additional TriQuint chips could jeopardize the DARPAprogram. Appendix B, Attachment 1 contains the TriQuint analysis of chipfailure problems; Attachment 2 presents Lockheed Sanders' memo concerningthese problems; and, Attachment 3 provides copies of the viewgraphs ofLockheed Sanders' Program Status Review Briefing and System Upgrade.

Likely candidates for upgrades with the improved AN/ALQ-126 wereidentified. Even if the Airborne Self-Protection Jammer (ASP]) is successfullycompleted, the AN/ALQ 126B upgrade need not compete with the ASPJ, sincethe Navy has declared it does not intend to retrofit ASPJ into the rest of the fleet.Mr. Butler noted that the administrative responsibility within NAVAIR haschanged. Oversight for the program now rests with Mr. A. C. McMullin, thedeputy director, Program Manager Air-272 (PMA-272), who may be moresympathetic to a flight test than Mr. William Brockner, the Deputy Director,PMA-253.

2.2 Subtask 3.1.2. Technical Studies and Assessments

Task Objectives. The objective of this subtask is to provide technicalexpertise for translating microelectronics performance benefits into operationalbenefits. This requires performing systems analyses and providing backgroundinformation on technical and programmatic aspects of a wide variety ofoperational and developmental military systems.

General Methodology. Key personnel used a variety of analyticalmethods to conduct technical studies and assessments. Expertise availablewithin the team of key personnel was supplemented as necessary by corporateresources and special consultants, in order to provide MTO with the most robustanalysis available.

2/1/93 MTO SETA Support: Quarterly Technical Report 3

Technical Results, Important Findings, and Conclusions. Theparagraphs below describe several study activities initiated under this subtaskand their associated findings.

Flexible, Intelligent Microelectronics Manufacturing (FIMM).Booz.Allen personnel continued to provide technical support to the efforts of theMTO Director and MMST Program Manager to create a small, flexible, intelligentwafer fabrication (or "fab") facility by the end of the decade. In this quarter,version 1.0 of the computer model of dynamic random access memory chip(DRAM) manufacturing costs was completed. This model was constructed frompast data (1970s, 1980s, and early 1990s), as well as current forecasts, to projectDRAM manufacturing costs out to 2004. Execution of the model shows capitalcosts for wafer fabs will escalate to $2.4 billion for the 4 Gigabit DRAM.Similarly, development costs for that technology node rise to $1 billion whilemanufacturing costs will increase to $1,958 per wafer. This model illustrates theimportance of increasing wafer diameter. Moving to 200 millimeters is essentialfor the 1990s and 300 millimeters for the 2000s. In the model, development costswere amortized over three wafer fabs under the assumption that a companywould spread manufacturing to multiple worldwide fabs. Even with theescalating costs shown in the model, however, the DRAM cost per bit continuesdown the historic learning curve.

Commercial Uses of Artificial Neural Network Technologies (ANNT).At the request of the ANNT Director, Booz-Allen initiated a study to documentcurrent commercial uses of artificial neural networks. Booz-Allen prepared aspecial technical report that is currently under review. It will be delivered to theANNT Director early in the next quarter.

Look Back at GaAs. The MTO Director requested an overall review ofDARPA's Digital GaAs Insertion programs. For the MTO Program Review inOctober, Booz-Allen assembled a retrospective briefing on the GaAs Insertionprogram for the MTO Director. This included retrieving historical presentationmaterial from Booz.Allen's files and querying GaAs contractors on their plans forfuture uses of the technology. The GaAs Program Manager adopted thisbriefing for his presentation at the final GaAs workshop on 20 November 1992.Appendix 2 contains a copy of this briefing.

Integrated Circuit (IC) Manufacturing Research. In February 1992,Booz.Allen initiated a concerted effort to study IC manufacturing processes andissues to provide the best possible support to members of MTO staff involved indesigning a new DARPA program for microelectronics manufacturing. To thisend, as described above, Mr. Connor attended that Microelectronics Processing'92 conference. He began a review of the proceedings of the latest Very Large-Scale Integration (VLSI) Multilevel Interconnect Conference (VMIC) conferenceas well; both tasks were undertaken at the direction of the MicroelectronicsManufacturing Science and Technology (MMST) Program Manager.

2113/93 MTO SETA Support: Quarterly Technical Report 4

mI

Booz.Allen believes that MTO would benefit greatly from a state-of-the-art

special technical report on IC processing, and recommends that such a task beestablished with a completion date of July 1993.

2.3 Subtask 3.1.3: Technical Documentation

Task Objectives. The objective of this subtask is to provide writing,editing, production, and technical document coordination.

General Methodology. The level of support required for this subtaskvaries according to the needs of the MTO Director and the program managers.Due to the rapid turnaround required for most documentation produced underthis subtask, Booz-Allen generally schedules all other less urgent tasks aroundthe generation of these documents.

Technical Results, Important Findings, and Conclusions. Technicaldocumentation generated this quarter included the production of view graphsand briefings. In addition, Booz-Allen personnel reviewed articles to besubmitted for publication, and the content of a Broad Agency Announcement(BAA) for a new MTO High Speed Circuit Design Program.

Viewgraphs and Briefings. A total of 60 new and revised viewgraphswere produced during the third quarter. Many of these taskings required rapidturnaround and absorbed a significant portion of available resources. A majorityof them were required for the MTO Program Review in late October. Booz-Allendelivered thirty bound copies of this review on 19 October 1992.

Booz.Allen also created briefings on the funding history and performanceof the Martin Marietta Radio Frequency (RF) Hellfire GaAs Digital SignalProcessor (DSP) program and the McDonnell Douglas GaAs Insertion OH-58DImage Processor program. Both these tasks were performed at the direction ofthe Digital GaAs Insertion Program Manager.

Article Reviews. At the MMST Program Manager's request, Booz.Allensummarized 14 articles that had been submitted to the program manager.Booz.Allen delivered these summaries to the MMST Program Manager, withrecommendations concerning which articles merited his further review. Manyof these articles were subsequently published in the September-October 1992issue of the TI Technical Tournal.

Editorial Assistance. Booz.Allen proofread a proposed MTO BroadAgency Announcement (BAA) for clarity of substance, as well as for errors of3 spelling, grammar, and punctuation.

2115/93 MTO SETA Support: Quarterly Technical Report 5

Is

Special Comments, Implications for Further Research. The level ofeffort required for this subtask had dropped during the last quarter, becausemost MTO personnel were at the DSRC in La Jolla, California, for most of July.During this quarter, however, demand increased and was particularly intensearound the time of the MTO Program Review, when all other tasks were deferreduntil the briefing was completed. Periods of high demand for these services areexpected to continue throughout the period of performance. To ensure rapidturnaround, Booz.Allen will continue the general approach of scheduling all

other, non-urgent taskings around those requiring rapid turnarounds.

1 3. Task 3.1 Management and Administrative Support Services

This task contains five subtasks:

* Program Planning and Controli Facilities and Logistics SupportI Documentation Management and Control* Administrative Program Support* Transition Plan/Scheduling/Phasing

3.1 Subtask 3.2.1: Program Planning and Control

Task Objectives. The objective of this subtask is to provide MTO withSplanning and control assistance. It includes (1) providing schedule and resourcevisibility to ensure that contract objectives are being achieved, and (2) trackingperformance, schedule, action items, and subcontracted efforts.

General Methodology. Activities in this subtask complement efforts inSubtask 3.1.1 (Engineering and Technical Integration and Coordination) and are

I carefully coordinated to ensure complete follow-through on action items.

Technical Results, and Important Findings and Conclusions. Duringthe third quarter of this contractual effort, Booz-Allen personnel performedvaluable program planning and oversight functions for MTO. Much of this workwas done for the FIMM, GaAs Insertion, and TOPS programs, as well as the newj High Speed Circuit Design Program.

FIMM Program. As previously noted, Booz.Allen personnel traveled toAlbuquerque, New Mexico to review the technical direction of the SEMATECH-funded CFM Research Center. A technical presentation was made at Sandia on"Cleaning of Silicon Wafers" emphasizing the need to work on all-dry cleaningand processing. Meetings were held in Washington and Dallas with Mr. ZacharyLemnios, Mr. Robert Blewer and Mr. Charles Gwyn of Sandia and Mr. Eugene

2/15/93 MTO SETA Support: Quarterly Technical Report 6

Feit of SEMATECH to move R&D to leading-edge technologies. Considerableprogress was made, and now Sandia has all-dry cleaning i anked as a top priorityprogram. Nevertheless, much more work is required to establish Sandia as aworld-class CFM research center.

GaAs Insertion Program. Mr. Butler attended a review at LockheedSanders in late August. Results of this trip are discussed under Subtask 3.1.1 andAppendix B. Booz-Allen arranged a classified meeting for 4 November 1992between the representatives of Lockheed Sanders -nd the DARPA GaAsI Insertion Program Manager, and Booz.Allen continues to monitor developmentsin the Lockheed Sanders insertion effort.

I High Speed Circuit Design Program. Booz.Allen staff assisted the GaAsprogram manager to begin a new program on high speed circuit design. Supportincluded rendering technical advice, assisting with the preparation of the (BAA),and suggesting potential program names. In addition, Booz-Allen compiled adatabase of names and addresses of individuals who were to receive copies ofthe BAA in early November.

TOPS Program. Booz.Aller personnel assembled a schedule of TOPSdemonstrations, informal reviews and IPRs to assist the TOPS Program Managerin planning his schedule over the next year. Appendix D contains a copy of thisschedule.I3.2 Subtask 3.2.2: Facilities and Logistics Support

STask Objectives. Under this task, Booz.Allen provides the peksonnel andresources to conduct unclassified and classified technical conferences andworkshops. Our support includes planning and organizing these meetings,mailing registration materials and proceedings, and providing conferencematerials and audio-visual equipment for these conferences.

SGeneral Methodology. Booz-Allen participates in various aspects ofconference planning, ranging from limited support to total support, asdetermined by the role of other participants and as directed by MTO staff.Planning activities include identifying, inviting, soliciting, and registeringparticipants; locating and pricing facilities to secure the optimal location throughan informal competitive bidding process; arranging logistics such as catering,equipment rental, conference set up; and providing follow-up as necessary (e.g.,generating notes and proceedings to document outcomes, thanking participants).

I Technical Results, and Important Findings and Conclusions. Thisquarter, key personnel planned the final GaAs Insertion Workshop scheduled for1 20 November 1992, in Reston, Virginia. Invitations were sent out to former

2/15/93 MTO SETA Support: Quarterly Technical Report 7

I

II attendees and other interested parties; the agenda was finalized; and all the

necessary meeting room, audiovisual, and catering arrangements were made.

I In addition, Booz.Allen has begun assisting the TOPS and PhotonicsProgram Manager with an Optics Review to be held from 8 - 12 February, 1993.Booz.Allen solicited bids, and identified the Hilton Hotel at Hilton Head, SouthCarolina as the proposed site. Work on this task will be conducted under theOptoelectronics contract, but part of the review will be of the TOPS program,which is supported as a part of this contract.

Booz.Allen also assisted Professor David Casasent of Carnegie-MellonUniversity in arranging an invitation-only session for TOPS contractors at the1993 SPIE OE/Aerospace Sensing Conference, which will be held in April inOrlando, Florida. On behalf of Professor Casasent, Booz-Allen solicitedinformation from the TOPS contractors who plan to make presentations at thesession

3.3 Subtask 3.2.3: Documentation Management and Control

Task Objectives. The objective of this subtask is to coordinate programdocumentation activities by building and maintaining accurate databases onprogrammatic information.

General Methodology. Taskings in this area are designed to supportprogram planning efforts covered in Subtasks 3.1.1 and 3.2.1 and provide MTOstaff with maximum visibility and flexibility to manage programs.

Technical Results, and Important Findings and Conclusions. Thisquarter continued to see important progress on the Financial Tracking System, aswell as several smaller databases utilized by the Microelectronics Manufacturing

I program manager in order to coordinate his programs.

Financial Tracking System. The main effort initiated under this subtaskis the design and development of the MTO Financial Tracking System, arelational database designed to track funds expended under all MTO contracts.The overall objective of this database is to provide MTO staff with betteroversight on the expenditure of funds so that MTO resources can be moreefficiently budgeted, committed, and obligated. The database will also enableMTO staff to identify potential problems or funding shortfalls before crisesdevelop. It will also register the receipt of contract deliverables and track thefinancial information included in such reports. When deliverables are late orfinancial data are either missing or incomplete, a letter report will be generatedand sent to the delinquent contractor, following approval by the responsibleMTO program manager.

2/15/93 MTO SETA Support: Quarterly Technical Report 8

I

I

During the third quarter, data collected through the Contractor FundingStatus Questionnaire was compared to data already collected and a program waswritten to merge this new information into the Financial Tracking System. Thisinformation makes the database much more robust. Dr. Andrew Yang'scontracts have been added into the system and reports on the status of hiscontracts are regularly delivered to him and the COTR. Much of Dr. BarbaraYoon's contract information has been entered into the system as well, andBooz.Allen staff will be meeting with Dr. Yoon's SETA contractor in order toprocure the data from their Contractor Funding Status Questionnaires.

Other Document Management Taskings. Other documentationmanagement and control taskings include continued, biweekly updating of apoint-of-contact database using FileMaker Pro software for the GaAs andmanufacturing programs; transferring a part of that database to the GaAsProgram Manager's portable Wizard organizer; and organizing the hard copyGaAs and manufacturing program files. Finally, as previously noticed,Booz-Allen compiled a database of individuals who were to receive a copy of theBAA 93-03 announcement for the GaAs program manager.

U3.4 Subtask 3.2.4.- Administrative Program Support

I Task Objectives. The objective of this subtask is to provideadministrative program support for the MTO Director, Deputy Director, and3 staff.

General Methodology. Many individual taskings in a broad range of3 areas were accomplished under this subtask.

Technical Results, and Important Findings and Conclusions.Administrative taskings performed during this reporting period includedmailing copies of MMST wafer displays to Air Force MMST personnel; arrangingfor, delivering, and hanging a display of the Martin Marietta GaAs On-BoardProcessor at DARPA; preparing a mailing list for announcements on a new HighSpeed Circuit Design Program (see 3.2.1); and obtaining various documents andinformation related to dual-use technologies and the defense and civilian

I portions of the federal budget.

3.5 Subtask 3.2.5: Transition Plan/Scheduling/Phasing

I Task Objectives. The objective of this subtask is to provide a detailedTransition Plan outlining the strategy and methodology for transitioning the

Sprogram support services performed under this contract to a follow-oncontractor.

2 /15/93 MTO SETA Support: Quarterly Technical Report 9

I

I

General Methodology, Technical Results, Important Findings andConclusions. The general methodology used to accomplish all taskings wasdesigned to ensure complete documentation of activities and development ofstand-alone, easily transferable deliverables (e.g., databases, special technicalreports, etc.). The objective in doing this is threefold: it provides MTO withcomplete records and thereby enhances the ability of MTO staff to manageprograms effectively; it causes the development of products and services that canbe completely transferred to MTO staff for direct use; and it ensures thattransitioning to a follow-on effort will be smooth and efficient.

I4. Conclusions

I This section summarizes the results of the last quarter of effort under thiscontract, notes any special problems, and makes recommendations for future3 work.

4.1 Special Comments

I Booz.Allen supplied a high level of support during the third quarter ofthe first year of the MTO SETA contract. No technical problems wereencountered in these support efforts, nor was any significant hardwaredeveloped or purchased.

This quarter continued to see significant contributions by Mr. GraydonLarrabee, special consultant to Booz.Allen. He responded to special taskingsfrom the MTO Director related to the FIMM program and reviewed the progressof related work by SEMATECH and Sandia National Laboratory. He alsocompleted the first version of a computer program modeling DRAMmanufacturing costs.

I Technical problems were noted in two programs being monitored as partof this contract. First, the poor performance of the first batch of TrQuint chipsfor DARPA's AN/ALQ-126B Upgrade GaAs Insertion Program has the potentialto jeopardize the program. Secondly, Graydon Larrabee noted thatSEMATECIHs preference for short term goals for the Sandia CFM is inopposition to DARPA's desires. As a result, Mr. Larrabee anticipates his role inproviding future direction for the Sandia CFM program will be minimal.

I 4.2 Implications for Further Research and Support

The computer model of DRAM manufacturing costs will be updated andextended to cover logic devices. Mr. Larrabee will travel to Asia to examine

2/15/93 MTO SETA Support: Quarterly Technical Report 10

I

Ifabrication facilities in Asian countries from 2 November to 16 November 1992.He will also attend the SEMATECH CFM Focus Technical Advisory Board(FTAB) meeting on behalf of the Microelectronics Manufacturing programmanager.

Booz.Allen personnel will attend the Government MicrocircuitApplications Conference (GOMAC '92) and will produce presentation materialsfor the FIMM Program Manager for that conference as well. The final DigitalGaAs Insertion Workshop will be held at Reston, Virginia, on 20 November 1992,and Booz-Allen will provide a broad level of support, including registeringattendees, overseeing room set-up and catering, and preparing and mailing outproceedings. Booz.Allen will continue to monitor the remaining digital GaAsInsertion projects at KOR Electronics, McDonnell Douglas, and LockheedSanders. Support will continue to be rendered to the TOPS program as required.This may include attending several reviews and/or demonstrations, as well asfacilitating the involvement of TOPS contractors in the Optics Review scheduledfor February.

I Administratively, the BAA on the High Speed Circuit Design Programwill be distributed to a mailing list of 150 names. A database will be establishedfor compiling the evaluators' comments for each preproposal. Similar work willbe done for the HBT program manager's new BAA as well. The FinancialTracking System will be extended to cover Dr. Yoon's contracts, and eventually3 expand to cover the contracts of other program managers at MTO.

As noted previously (in Subtask 3.1.2), Booz.Allen believes that MTOwould benefit from a Special Technical Report on the current state-of-the-art inIC processing. It recommends that such a task be established, to be completed inthe third quarter of 1993.

2IIUII

12/15/93 MTO SETA Support: Quarterly Tedhnical Repout 1

I

IIIIIIIIUi APPENDIX A

•TRIP REPORTSAN JOSE, CALIFORNIAI 21-25 SEPTEMBER 1992

IIIIIUII

ITrip Report

MICROELECTRONIC PROCESSING 92San Jose, California

September 21-25, 1992

Mark P. Connor attended the convention at the request of Mr. Zachary Lemnios,3 DARPA-MTO. The convention consisted of numerous conferences, several short courses,

and a technical exhibition. Of these Mr. Connor attended the following:3 - Advanced Techniques for Integrated Circuit Processing II Conference,

* Rapid Tlmal and Laser Processing Conference, and* the Technical Exhibition.

The Advanced Techniques for Integrated Circuit Processing H Conference lastedthree days and covered the following topics:

3 polysilicon dry etching,* plasma processing of materials,

3 - plasma process modeling,

- high density plasma sources,

- integrated sensors and control strategies,

- multichamber processing and cluster systems, and* compound semiconductor device processing.

Several highlights of this conference are presented below. Etching issues,especially with magnetically enhanced reactive ion etching (MERIE), dominated nearly half

of the conference. Of the many issues discussed, the increasing problem of microtrenching

warranted additional papers. See figure 1 for an actual example of microtrenching.IIIIII1I

UUp so 04 - 0.4 -L

I Aspect Ratio= 3:1.

I 1.2$ Polysilicon Figure is not drawn to scale.

1 =830 Gate Oxide

1 0 4 1

3 0 0 A* S SI , Mi rtrenches

UFigure 1. Microtrenching Example.

Microtrenching degrades electrical performance and exacerbates reliability

problems. Characteristically, it only is seen between dense or adjacent isolated lines.Moreover, it follows the grain structure of the polysilicon.

Stanford has been investigating the formation of microtrenching, and has

established a multiphase, plasma etch model, where understanding intrastructure transport

phenomenon, i.e. between lines, is critical. In general, microtrenching is formed by ionreflection from sidewalls, and, to some degree, surface charges. The proposed solutions tothe problem include:

(1) reduce the aspect ratio from 3:1 through the use of thin resists on the order of

0.55 g.m;

(2) reduce the etch rate of the gate dielectric, possibly using bromine instead ofchlorine; and,

(3) collimate the impinging ions with tighter tolerances.A general consensus throughout the conference was the trend towards the use of the3 DESIRE process for dry etching. The use of 02 plasma to remove the photoresist is

desired, even though it leaves a sidewall residue. Nevertheless, SiO3/02 plasmas havebeen demonstrated to perfectly transfer a pattern at 0.275 pm with no dimensional loss.Moreover, such plasmas do not need sidewall passivation.

Some discussion involved multi-chip modules (MCMs). Specifically, the cost

drivers in order of priority were identified:

*2

I

1. quality of ICs - cost driver for MCMs;2. IC testing and bum-in, and limited IC testing standards for MCMs;

-- e 3. ceramic MCM package - 99% are made in Japan which takes -12 weeks toI-- receive;

4. substrate; and,

5. assembly techniques, e.g. flip chip.

The market for MCMs appears to be developing as"* workstations,

"* laptop PCs,

- high performance PCs,

"* portable telecommunications,

"" machines incorporating an asynchronous data and voice transmission mode with

U high bit rate,

* cross point switches (10OMbit/s - 2.4 Gbit/s),I - DSP applications.

To make MCMs more cost competitive with continually evolving IC technology, one must3 identify high priority applications (subsystems) for which ICs could not reach the same

specifications in a reasonable time frame.

3 Now, consider the following figure:

I3

I

1 3

i

IU 0.1 - - -

E 4Air

0.5 1.0

Size of Defect Qun)

SbFigure 2. Defect Density vs. Defect Size.

-- Curve A in Figure 2 is the current situation in MCM manufacturing. Curve B is the desired

-- situation which would eventually eliminate all 1.0 pmn defects. This transformation's

m ~criticality increases with increasing substrate arma Similar curves and transfo,'mations existsam_ for IC manufacturing, but with different scales. The way to achieve this transformation

cost effectively in either manufacturing environment is through the use of

mcroenvironmonts. This is most facilitated with multichamnber processing and cluster

systems.

---- Another interesting highlight was the discussion about laser cutting and writing for

-- MCM customization. Such a process allows all but the two interconnect layers to be fully

-- customized and tested from a semi-customized intecnnc strctr. The x-y signal layers

_ consist of parallel 2 nun line segments which open up to the surface through a via. Thus, a

bridge can be made; or, if a bridge exists, an open can be formed. With laser writing ori cutting these via jumps, fully customized interconnect structures can be made, as well as

low cost repair. Since cutting is quicker than writing, one approach to speed throughput ofI such MCMs in a high volume manufacturing environment, would be to supply the semi-

custom interconnect structure with all the via jumps bridged. Thus, customnization resultsonly from laser cutting.

* 4

I

Yields from this approach of 99.9% have been shown with IC compatibility. In

fact, a demonstration model involved the Motorola 88100 and two 88200s (f=20 MHz).This I" x 2" Cu-polysilicon substrate contained three layers with 216 signal interconnect

nets. 4000 links were made measuring 20 mil ± 7 mQ; and, 1600 cuts were mademeasuring - 1010 LI This MCM safely operated up to 100 MHz, based on wavefront

dispersion. Moreover, following the initial writing and cutting, no further rework wasrequired.

The actual Proceedings of this conference are forthcoming. March 1993 is theexpected month of publication.

I

IPEDCBATAHMNI

IRQITAAYINAHAMETN

28AGST19

An Analysis of failuresin the S097]2A- D•IF

Api-rure Buffer

i byDavid Perkins

llj .TriQuint Semiconductor, Inc.

" '~forS~Ed Wassung

, Lockheed - Sanders

Fy~x j &o r s •~I

I1I

II

Iid.Lp. •~A An ly of failures 192st1139A

inteSi2 IM

Circuit Description.The DIFM1 - S0972A Aperture Buffer circuit consists of six 24 stage shiftI registers accepting data from an on-chip six bit read only memory(ROM).The shift registers are clocked by a clock source consisting of a MUXwhich selects one of two clock signals, CLK and CLYJ'2. The CLK signal isselected when the select line is forced HIGH by the negative transition of theCLK signal. Two CLK cycles later the select line HIGH reacaes the resetpin of the CLK/2 flip-flop forcing the CLK/2 output LOW. When CLK/2 isselected, the select line is forced LOW by the negative transition of the CLKsignal. Two CLK cycles later the select line LOW reaches the reset pin ofthe CLK/2 flip-flop releasing the CLK/2 output to toggle. Control is givenback to the CLK signal inunediately after the selection signal is forcedHIGH by a negative going CLK edge. Two CLK cycles later the CLK/2signal is forced LOW by a HIGH to the reset pin of the CLK/2 flip-flop.The MUX drives two clock buffers which in turn drive six more clockbuffers. Each clock buffer at the end of this clock tree drives twelve flip-flop stages in one of the six shift registers. For each register the clock driverorder is as follows:buff-I drives 1,4,5,8,9,12,13,16,17,20,21 and 24 for a total of twelve;buff_.2 drives 2,3,6,7,10,11,14,15,18,19,22 and 23 for a total of twelve.

Operating conditions.CLK is 825 MHz (a 1.212ns period). CLK/2 is 412.5 MHz (a 2.424nsperiod).The circuit has been set up such that CLK/2 is always feeding theaperture buffer. At a rate reater than 300KHz, CLK is selected for 200nsand data from the ROM to provide shift register loading ofalternating ones and zeros. After 200ns CLK/2 is again selected and the datais shifted out of the registers, i.e. the signal CLK/'2 is selected fo"20sýjThe signal CLK is then selected for 200ns and data is loaded from the ROMinto the shift registers. ( During the 200ns CLK window CLK/2 is heldLOW by the selection signal. ) The signal CLK/2 is then rselected and datais shifted out of the registers.

CLK CuK/ CuK CLWselectSd selectd selected SS!O~8d

200n, 200us 2C0ns

CLK - 1.212 no Penod CLK/2 . 2.424 ns period

The CLK - CJM timing relatonship when filures ae ob=vc&

dJ.p. Psge 2 of 4 11 Aug=i. 1992 at 11:39 AM

Behavior observed.When the shift-in shift-out rate is maintained at greater than 300KHz, thecircuit behaves correctly, i.e. a pattern of alternating ones and zeros is seenat the output of the shift registers (see attached plot-1 - jul 24 92 - 12:25:49 -rate is 500KHz). As the shift-in shift-out rate decreases, errors begin toshow up in the shift register outputs (see attached plot-2 -jul 24 92 -12:30:51 and plot-3 - jul 24 92 - 12:43:45). The number of shift registerstages in error is less when the last sampled bit is a one (plot-3) and greaterwhen the last sampled bit is a zero (plot-2). Except in those cases whereerrors occur in the first or last shift register stage, good data is present instages both before and after the stage in error.From plot-2, trace(tr) 1, errors are in stages 5,13,17 and 21. From tr2, errorsare in stages 1,9,13,17 and 21. From tr3, errors are in stages 9, 13, 15,17and 23. From tr4, errors are in stages 1 and 13. From tr5, errors are instages 5,13 and 21.From plot-3, trl, errors are in stages 5 and 13. From tr2, errors are in stages9 and 13. From tr3, errors are in stage I only. From tr4, errors are in stages9 and 13. Trace 5 is functional.Other data from aug 3 92 shows error patterns as follows:from plot-1 - trl shows stages 5,13,17,19,21 in error;, tr2 shows stages1,9,13,17,21 in error, tr3 shows stages 9,13,15,17,21,23 in error, tr4 showsstages 1,13,17,21 in error, tr5 shows stages 5,13,21 in error, from plot-2 - trIshows stages 5,13 in error, tr2 shows stages 9,13 in error, tr3 shows stage Iin error, tr4 shows stages 9,13, in error, trS is functional.Daia taken of the shift registers, prior to the CLK/2 signal selection, showsthat all data is valid as it progresses under the control of the CLK signal(plot-4). It is only after the CL=/2 signal selection takes place thaterroneous data is observed. Errors are occurring in shift register stages thatare driven by independant buffers in the clock tree. Errors are occurring inshift register groups 1-3 and 4-6 which are driven by independant buffers inthe clock tree. Erroneous shift register output appears to be related to:1 - the rate at which data is shifted in and out of the registers;2 - temperature, ie. when the part is cooled, fewer errors are observed;3 - (to some extent) the value of the data in the register stage at the timedata loss occurred ( although errors have been observed when data valueswere either one or zero); and4 - frequency, i.e. at fctx = 800 MHz and selection rate < 300 KHz or

at fca = 10 MHz and selection rate < 3.7 KHz.The clock circuitry ( as observable from the SDLSYNC output, plot-5)seems to be functioning as expected ( it must be understood that there issubstantial filtering of the clock signal at this output because of the numberof cells in the chain and the bandwidth of the output cell which swamp out

d.Lp. Page 3 of 4 11 August. 1992 at 11:39 AM

the observability of possible glitch effects).

Analysis.Since good data is being shiftcd c'it (except in the cases where data is lost in

the first or last stages) before and after erroneous data, the stage which lost

its data must have been functioning correctly both before and after the initial

data loss in order for prior and subsequent data to exit the shift register

correctly. The data loss must therefore have had to occur at one and only

one point in time, after which functionality was restored, in order for prior

and subsequent data to be valid. Since data loss is happening to more than

one stage in more than one shift register path, the cause must be related to an

event which is common to all shift register stages in all shift register paths.

Given that all stages are functional during some portion of the shift-in shift-

out process, and given that the ciruit function is normal above a certain shift-

in shift-out rate, we may conclude that the most likely explanation for dataloss in individual stages is that some kind of clock event occurred to causethe data at the input to the stage to -p-ass prematurely to the output of thestage. This event must have happened in a clock path which is common toall shift register stages since data loss is observed in all shift register paths.The only device common to all shift register stage clock inputs is theSF2MUX cell which provides for the selection of the CLK and CLK/2signals. Also, since data under the control of the CLK signal is always goodprior to the CLK/2 selection we can say with certainty that the clock eventcatising the data loss is the result of either the CLK-to-CLK/2 selection orsome fundamental problem with the CLK/2 signal itself. Now, while theCLK/2 signal is selected, no current is flowing through the source coupledFET pair on the CLK side of the SF2MUX, even though the CLK signal ispresent at the gates of those devices. During the 200ns CLK selectionperiod, alternating one and zero data is shifted into the registers and appearsat the output of those registers without error. During this time the CLK/2

v signal is forced to a LOW state and no current is flowing through the sourcecoupled FET pair on the CLK/2 side of the SF2MUX.

I Some questions that immediately surface are:Sj 1- what is happening in the SF2MUX that could allow for a clock

disturbance such that data in the shift registers is affected?; and72 - what influence does the MUX selection signal have on the output of the

, NJUX, such that a clock disturbance could be possible?

.A ,Y:

d.Lp. P-AgC 4 of 4 11 AUgust. 1992 at 11:39 A.N

IA 602 DIGITIZING SIGNAL ANALYZERI ate: 24-JUL-92 time: '12: 25: 49

1APERATURE BUFFER UP71,250omy - I - 11

A A. -A SI~~

11 Li

/dov = ... . _

I i V Q

, .Utizity 2 ,lW'm 4.

LI

Intalz ,-ii - bl e •Text

i aite Vert Pas . ,Horz Pasl1 OfIfi n-

/.....dP,•~,._l.•si~•v O..I...

Ipor

I.4- )SA 602 DI'GITIZING SIGNAL ANALYZESo

Idatte: 214-JUL-922 time: 12: 30: 51

-25MY;APE.SATURESUFEA OUTPUT

TiII rI7

N AvIdlyTA

IToI

I trlg'a%

IwMA I4

-2. Pns Time I.ns/d0 aae ooesPoes Clrta Hc zOM

1i Pfm I - ..I

,SA 602 DIGITIZING SIGNAL ANALYZERdate: 24-JUL-92 time: 12: 43: 45

APESATURE BUFFER OUTPUT-250m-;

i" I

* S

I "rlII

1ivtv-,ve

I~~0 -!, "fi--iu• ... , --" ' - - - � - - .....

"025 'I " - -'----

052-e C 04%oar F; 0.emjutiý' p' 2 'Wfm 41

' Li

iti~eT.ne Lan i ext e Tex,*Oat- iar_.o Pns fHarz .os

12. 43: 45 *Oisp: Of f~0 0sQ..

"OSA 602 DIGITIZING SIGNAL ANALYZER* date: 21-JUL-92 time: 44: 40:03

II.' 200MnV

/dlv

2 -- .5V..3 -14ne 5Ons/div 486ns

-1-M

200mv3 Idiy

I II -2.4.5V II3.n ln/~ 438. 7ns.

TrI gge, Source Level Time ModeIS;Wlect Desc Holdoff ;Wfm 2

Main Li i3 7sjs Auto ILI __-_-

'Coupling slope Timer tiTimer t2

2lv 15+ n

I3.n Qsdv487s

OSAM 602 MIG:7IuING IGiNAL ANALYZER3 date: 10-AUG-92 tvne: 16: 49: 32

41 r~ a" - -

AM ~AA

),~ a -va

I .2ne iOns/div E!i i2n

-4=1 -

trill

* R2

-3.* 5V3 63.O06ns Op/4v - 68.0~n~sSP'3 RS '2 2-ýC a- Hr. dca.p-y- Remn

T Ta ýtk Qny-' 9600 baud HPGL.-LF.xtifdecl Self -7eksecu.re Tmain T Jgai a ~f Time

'Oiag~ncs t -c. 7etM e e i Tr~ig H.ldOf f-900m/ 2.us I

I

IIIIII

IAPPENDIX B,

AIrACHMENT 2LOCKHEED SANDERS ANALYSIS

NASHUA MEETING28 AUGUST 1992

III

III

I

S ',?Lockheed Sanders, Inc Memorandum

To: Distribution File Reference:

From: E. Hollis EH/eh/1 -2113/92-28

SubJ: A Few Observations, Suggestions, and Date:August 26, 1992about Triquint Fixes

Ref to two previous memos by the writer plus fax from DFP of Triquint

My Suggested Reason

As noted in one of my previous memos, I suspect that a lowered E mode devicethreshhold possibly operating in conjunction with sidegating due to manufacturingtolerances not being adequate and also in conjunction with hysteresis may be the culprit.

TrIquint's Hypothesis (They know their product better than I do.)

Triquint thinks the problem relates to charge storage in the crystalline defects. Theirsolution is have a small amount of current flowing through the top of the off differentalpair in the top of the cascode network which makes up the MUX which selects the clockor clock/2 signal. I have crudely drawn in the figure below what I believe they are tellingme. The upper right side of the differential pair is the 'off' pair in this example and wouldrepresent, for example, clock/2. Without their fix, the current through either member ofthis differential pair cannot exist because the lower part of the cascode is open circuited(has not been selected by the select line which is in effect DSTOPBIT.

Triquint's Modification to the MUX Macro

They intend to modify the MUX by adding a separate path -Vee to allow a small amountof current to flow even when that side is in the 'off' state.

My Concerns

Their theory is that this small amount of current flowing through the differential pair whichis in the off state will overcome the problems arising from the crystalline defects.

I have two concerns about this modification. (But again, they know their product betterthan I do.) First I don't like the idea of feeding current through one or both sides ofdifferential pair which is set up to be regenerative. Secondly, I especially don't like thiswhen I suspect a degraded noise margin due to a suspected lowered threshhold of theE mode device and/or sidegating and hysteresis. My overriding concern is that while thismay cure the problem in the lab, it may become a disaster in real life operation. (By the

Page 2U. EH/eh/1 -2113/92-28

I same token, we'll never get to real life operation if we don't allow the vendor to do whatit thinks is best and if we keep raising too many problems whose probabilities of

-- occurrence are not known.)

E. HollisProject Manager Custom VLSIMicrowave Technology Center

i

Distribution:

E. BachelorF. DeCaroJ. Sylvain

I E. Wassung

i

Iw

IF

i

I

Ik .

5I

I E. HOLLIS 8/19/92

I ORRIN RESISTORS RNO SOURCE FOLLOWERS. ETCARE NOT SHOWN AND RRE NOT NECESSART FOR THISI EXAMPLE

I!

• m m Iw z -

Ii

APPENDIX B,|ATTACHMENT 3I

LOCKHEED SANDERS BRIEFING NOTES!NASHUA MEETING

28 AUGUST 1992

I!

III

1i

III

I

Clo

> M-4U

+ 000CCo

z

00 0L

wl *#MPL

a; ~

z CLU M

0. F >.

LU 40 5 8

LU L

m0 CL

I U0,-00

LU

I-.

62

0_- I- I0.-mi

., - .... ...... . . z

0 0

c c L uI S2

..... .

A

'I

-A *� � it

* -- A

* z01

C, Lu--

IIU-LU

I 0

....... ...... ............ ...

IC

0.0.

CMo

IIII. -- ...7--

- .vt

U I

I..I.I1:......

- £ -- *�, *y;�* cI

UIUI* $

IIII

III*II.III*UIIIIUIII

j 0

w 9

z

0 06

I InI- z

0 cc

z 0

I LU a,'.

ulz -0z

ClO)

zz

ijuIU a

IiiUL

0

1WUlULUuj LU

z m -LU

Lw L

*j 0 1- m *-A. LU LU

Cc 0 LU Zu' 0 rim Cfl 2

LU 00U, -

K WLUJ 00

0) o* ~u LU U, W W 00 o

m. 0n

3:! LU NULU 0

2ý 00 LU

0i 0Lu 0

0 LI-LU 0L

LU L)U,.

Imu

0 o

' --C"- I

3;1 I~fR\ oI. h I S .

NIP;'I~~t.SU~~t__

___ _ __ _

n* i ii

0. I.001_1 I-

fl If

LVF

S

--

"I : a , I, !.. . . . I I

- I * F

b|. *I•.

0Iaz

z z1 0

0 0C

* 0

0

zU WIc LL W w w

Ii 0

L j6c CL 10 zml0 0

ccE

IL OWIL 0 ~ 0

FE I!MI 511

I ~ g L06

ccI~U6 cuLL

vi W.?

IY U

- Bic

3L cc---

I2

IU Co t

S LU I-L

COco

a 0c

-0 a, 00 0uU

OW1 I-CD LL..

IU '(L

CW C.)

Ic

- 0 V

%ito IL ZAmp

C,, 0

I.0 x0

Cl)4

z a zz iLU W W O)4

0. I-

moa zu- CCc4LU L

Ii LL CJ0 C.)

ij W(.).cmoL oWz OC-- 2 ml 0~

0 0 6 * *0

saw.

1L~IL

jj

-t.

f � �. -

-. � -�

-p1

$

i.-' .- �'.

ra-

-4-'4-.

- * -.-- �.** -

�

4 *t, .0

-r -

-4' -

7'

*t. S

*a-.

ft

'. 4w�-

�

��# p., '±,

V <S -

* *1*�j2t

a -

4'--- S

S --' Ld

-alN.

a� -- - -.

S -.

U-

La.

It'

(

4"

z

2

I-. w C-

o W

=CO) U) I

0 1 L CF a

>. cc 0 z(I) Lu u 0 t-LU

>~) InL

= ILI

I II L

Iz

3 0

Vcc CA)

5 0U LU CC

o. z

CC 0 0~0 OC .ZCLwL .og

CCIMa.c.c 00 C0 UJ L1UU 0 uJLULOz .

ij j C 0 0U

LU CL 0 .4 (R -rUc O

IiC U rI U 0u

ULuJ

LU 0

0 0L0Om~cc WU

U) w

3U 00. W a w.

ii( C. 0LU~2 4 F

I- cc 02Ul) dCC

0 0 =a jib0

0 >

0 CL 0 0

3r

cc z1 0.Ia w I

0. .

I Ia C ,ILU zo0

jz IIS0 us a9 UJ

I S wzLUL ZC

I- ra w ~r~ ,

COz CO .zo0cr 00 0 0 L

Iz W 0u* 0

00

IM~h CJ~m 0c

4 0~ 0o 0 49 c t

0 w w

a. U)."'1 a.o U

Im- POZ z I-a

a. I' Ic0. cc

U Iw C.-

us a o0.

U..

d 4 a

lo 0

m crw C 0 ZLU 01- 0.

49 )

LU -- ..1 L!) Cc

0 0

I L

I za. 1LLILLI! a1 0

m FU

oL 0 *0

zF

0~4 a.a

o 0z>

C I0

Ii 0)01U. 4

0 a

LIU

uLu

zozR LA. 0 1

*i - 0 0Z' Lu - 0 A l.

C) .1 C ,IL Luz30 0 U. .040 o. 0l

cn LL -JL 2L3j >1i 0~I.i a 0

u l ) C..) Z IU

z z 0 U004

IiiCIM

Iz

* z w

Urn 0 Lu

Urn ZI0 LU Ur I

0 _ L wiz

LU0 O

0.z wi w1o0 4t I-- a0- 4 C.)

0

00z cc0U w

4 ILl

3 0 0 0 0

IL

a t)

77

a% -a

W 0

Z OC 5au' CL W

4c wool0 c C 2c

4~~0 cmCCU ZL

0

0R

CD5cU z1~2

o I-

4 4 0

um. z0U 0 0

-4 >0 0. C5

4 wu U U MA

0 Oj I c( U c

U. 0~, 1 ,0. CL cr

LU

IL.U

0

z Ro D wo COCM to fn L j0

z 0

cCJ J L 3Co O'C 0 oCC Z L

LU wU0 c c5J z _0.0

tz Lu

z 4Lcr. L

cc0 0 0 0-

40

LU UrL

S 0FLU a

0 a OZ =. !a

On( LU0.jco ~L l 4c I-r

C.) 0 40 P0

LU 0-1LUC.,

CL ii. iii L. LZ C~40 0 0 z

=) 0) L 0 LzL LU0 5 I-I.-l IL L

z ~ .2 LULU 0Z X 0

0 a L0L 0 z0 00cE.LLU Z UQI

5l LUQO. 2

IIm LU Z ca

mu 0 0

CC CC * ) 0 0 0 cz

4400

0 4c4CCI ( o )C

4U L .Co i 3!3:a ZL. L-

04

LUUcc

C.))

-J a

0

u.J 1 U.1-0 'U

0 0 Cl

I cmUo

MOO 0 c

cCLIL 1-zL0 P a

LIJa

Wa

Ew

Cc CoQ ZCD*W(d W

Z ý- *

wzz

CA c)V zW W i a 0

LU 0 0 E~ F. ocSN4

CJ 00

z Oujj4O WEU

0U mm 1z 0 > c0P I

D-JV U. U 0 z c

Co O5tn 0 0L0im-I N l

00 WOW Ul LL. j

a. 6 . cc 0 > U.4c LuZ U C i-

* CA z 0

cc JL = = z4c0

I z

LU

I z

zwLU ZL

LU L-2

In Z0 0 (Lfri O

C* 0.0

U.. 0 t;00.4 >*0 . 9 0U

Cc~- 0 c

Im000I(D Z0 LL U UIc OC a)

I 004

I UC

* 0LI-rP

0 0 Z .

z w L

z cc a

CC U .Z 0 w

0. 4c C.w LL .1 -

Z ~ u 00W

z 21-0 La OC J

t~o IL1 0.zz uJ)Cz 1O wLLL Zww J

IL + 22~-.IL 2

z OWE 0) 4

z 0.3 0 Wg*

0 0z 0z z c04 4ccc t; x Lx z z L-

u W WUL 0

cc LU

LU Q.

0 LULU LU.I A z

w j0zL LU0 0) 0.

-L 00I O o LLUI U-

Cl) ~~ U)z2- LpU LU

= ai.I zcc LLI L

0

0 I-0

ILL

00

II

za

I - - - - - - - - . - -

I - - - - - - - - - -

I 1-�--��� - -

a

II ------ 4=---I --- 4----"=1I �I

wU.'-

I a - - - - - - -

Cl) I -� - -� - -

I I.

I 'S

II Ci) 2

- iii1 -w

* ('3U - - --- --- - -

II

Im

z0 I4

LU 0

4 U 0

3- z0 a m

z~ CLIL-aA 0 x

LU o4ol

II- LU0 0U L

C0 C 0 a

Z0 z ! t 0. 0ZU (LL L~ Fo LL.V

0 LOU L LU

0 0- 0 0 ccml z 2

o- zi u

LL. LLw 0LL,%0

0D LUU c

0 LUI-

0 0

U. F U CoLU me mJ L

tI- FE

LUU LU UOLU 2 +-1-> IL

I-. 9L cr. Z C

o U ILWLLULU 0-

r) LU a 1

CoCO) Z 00Cl) LU 1-ul

LU LU -"1z CL cc >0

W0 CC

IL 00 > UOLU w0 zCDUL z I

w(2 (cri LU U1

+ U. F Z

CCLU 0 0.ui IL C z c1j~ ~ 4 mJ -LUlL

LUc Z2zcr

I

m APPENDIX C- BRIEFING ON

DIGITAL GaAs INSERTIONS:A REMROSPECITVE20 NOVEMBER 1992

0

wU 0Ir BE-

E

I ~z

0I CU

wI ~Cn C

zN

cc

0Ato

ww

IC.)'LL

ICD) 04 L

co 0o

w

Il m

0 MMROIM 01 0Cu mVmc

-JW

crZ

(1)0 u.

o0b > O.EA

m(01- z LLI

(10 0 CO)-4

-0 4cj

a. 0~ Ul

L)0

(D 2 zLU0

LL -!

0>0 0 1-.nz 5LU z

0 0 C ý0

on 0 w~ w

m z i. F .Iin IC.

LU

uw 0 c CD 0 L0n 0 0. -2

NOCLW Z M mL20 C ný m a0 x w

..000 .m~~ Ui 0 LUO~

-J 0a. OUoCO

Z - M W

LULU -j ix oz;0 J) 0LU -L AL

U. cnmI Wil 0 . LU0 0 0 c 0

0

- 0

ILI.

32~3 4 .aam

41 30 00 1W c i i6 0

IL 4 6 IL 'Z. 2 0 -UI

vi 2 I0 i- x p m p3P. a m1 n'

0 S M

a 41C

ILc 0 Za c a So,-A U. u s.- 3 * a

m O- a z oPw0..

4D >- c 0 W, tI

00 au= M a p0 ..cc O ru a =. no ,za as: a. . is4 = - t

o0 (1 ~ a 30 w~ w- 0 cP-u0IC ~ .. ~gI- m .. 4.ac .. E tp P4

c16 TO 0 =

P-4A .a

Z rn 0 cc 0I z-z z4 Ac acc

2 4c Ow Il $a. Sb

LL,1 4.a C3 =5- C4 4U3 ýa

a.. W IL d.EAu IL 4A 0 Su qt t = i- -c .w 4

ILI 3c - I.-L

__ 0. a0 4 Z=LU U-C ~ S

C) ** .4 wi aý a- 3-03 0 a ki 3w a0 - 3W E IL COU 20 C 4C S0 300 1 C 2

IC0 0U) IL cL

i W-c enC 0lI- ca a) wi C A u... 0 2 0 IL 00 w 0 0 0 0 0.

a 0 0 =W6 ~OIL 0 zc a0 0 a3

a i aCU,45- 44 a

a- W 0

2 0 41a a

W W41 W~ U (A 3Z 07

a ~0

-U -C -C - -

I-cc .c 00 0

(Uo 0' am94OR c E c'aL mw0 Ch cLSm emEDem0)0) CU*0 .

COG .v 08000

coC 0 h. 5 0 0)C

0 ROE WEo ~ ~ ~ CD 0 0c 0

_w Im-Ila0 00E

CO)~ CLUS

CC C 3: 0_ C

0- cc0U0 > SEM

Immm 2o-0 ~~S00

I- 0 i. ~EGOWN iom

M C0 mU) mU~. CUm

C= 0mU U) ) U)

LOW M mmo

OccUW 0L cCU'GLuI( UUG U) 21 0 'm' U)c (Ao

ow1 0 CU 2I C E ccCDC?.. c ="~' CU & 4-

on 0 imUma 0- v CU3U)0-a r-c' 0)zeL

Mo Lu~c zS 'r CcWOMEN~ ~ .CL91lom WMM M >

"M 0~- CL c0 Iicm~ 0 ZCI 000oj

20 m 0 0 SLM

0

0 1 c0 I3

o 0U 000 M) CLZý 0 0 0.o L) w 0 2 c cLUM n0. 0. v 0 C

W.. 0 0 0 (DCh sm *0 O4w4. IM .2

LL~a 00)& > 0*zfla) CL eD IT4 4) o-Cc a .AO w %e 0

LL 0 a ac O- (A 96

I0 000 ) 00 C1 (A~0c 0 0U0) C O Cc *a C cO S .(JC C1 4 s CLz 0w cc) a)0

a 0 (A 6%4~a )~ 0El

(A co 0 0~"0 0~ .

0 6 0 we (a

CL

IIIIIIIII

3 APPENDIX DSCHEDULE OF

REVIEWS AND DEMONSTRATIONS FOR THE3TRANSITION OF OPTICAL PROCESSORS TO SYSTEMS(TOPS) PROGRAM

OCTOBER 1992-DECEMBER 1994

IIIIIII

I TOPS Schedule of Reviews and Demos

NoegI. Harris has stopped work on the EWC/PDF contract. Harris and NRL aremeeting this week to work out a solution. Future review and demo datesfor EWC/PDF are therefore TBD.

Harris OAM Dan Beard "informal" review w/ Rome(Melbourne) 407-727-6176 26,27 October

Hughes Phased Arrays J. J. Lee "informal" review w/ N.(Fullerton) 714-732-6828 Bernstein, 28 OctINovember 1992

I TI Phased Arrays Chris Hemmi semi-annual review(Dallas) 214-575-6269 (early Nov.)

I GE Null Steering Dan Friedman quarterly review(Syracuse) 315-456-4952

I TBE Pattern Recogn. Charles Hester review(Huntsville) 205-726-1999

3 BDM SAR Mike Haney demo(McLean) 703-848-5854IDecember 1992

3 Nothing scheduled.

Martin Mar. Pattern Recogn. Scott Lindell PDR(Denver) 303-971-3253

Dynetics Pulse Compress. Neil Mohon PDR; demo 1 -D correlator3 (Huntsville) 205-922-9230 (25-29 Jan 93)

Hughes Phased Arrays J.J. Lee "informal" review w/ N.3 (Fullerton) 714-732-6828 Bernstein

II

October 20, 19921I

* TOPS Schedule of Reviews and Demos

SFebruary 1993

OPTICS REVIEW 8-12 Feb 93 Hilton Inn, Hilton Head Island, SC

ERIM/ SAR Ivan Cindrich system reviewBDM Mike Haney (1st week of Feb.)TBE Pattern Recog. Charles Hester final build

(Huntsville) 205-726-1999IMa~rch 1993

I GE Null Steering Dan Friedman quarterly review(Syracuse) 315-456-4952 (Mar/Apr 93)

U Hughes Phased Arrays J. J. Lee CDR, lab demo possible(Fullerton) 714-732-6828 (Mar/Apr 93)

I3 SPIErTOPS 12 April 93 Marriott World Center, Orlando, FL

ITI Phased Arrays Chris Hemmi semi-annual review(Dallas) 214-575-6269 (early May)

* June 1993

Martin Mar. Pattern Recog. Scott Lindell Flyable Prototype(Denver) 303-971-3253 Environ. Eval. (at MM)

I -GE Null Steering Dan Friedman quarterly review(Syracuse) 315-456-4952 perhaps lab demos

IIU

October 20, 1992 2I

TOPS Schedule of Reviews and Demos

3 August 199Martin Mar. Pattern Recogn. Scott Lindell CDRI (Denver) 303-971-3253

m Dynetics Pulse Compress. Neil Mohon review; demo 2-D(Huntsville) 205-922-9230 correlator

ERIM/ SAR Ivan Cindrich CDR_ BDM Mike Haney (1st week of Aug.)

Se1tember 1993

Nothing scheduled.

TI Phased Arrays Chris Hemmi semi-annual review, demo(Dallas) 214-575-6269 prototype chip

(Oct/Nov 93)

November 1993

GE Null Steering Dan Friedman quarterly review;(Syracuse) 315-456-4952 possible processor demo

Martin Mar. Pattern Recogn. Scott Undell SPOTR delivery(at MICOM) 303-971-3253

December 1993

Nothing scheduled.

Hughes Phased Arrays J. J. Lee system demo(Fullerton) 714-732-6828 (sometime early 94)

Martin Mar. Pattern Recogn. Scott Undell tower test(at MICOM) 303-971-3253

October 20, 1992 3

-- TOPS Schedule of Reviews and Demos

IFebruary 1994

Martin Mar. Pattern Recogn. Scott Undell flight test(at MICOM) 303-971-3253

Dynetics Pulse Compress. Neil Mohon review, integrated system(Huntsville) 205-922-9230 demo

U -

Nothing scheduled.

Amil12

y Dynetics Pulse Compress. Neil Mohon MICOM radar demo(at MICOM) 205-922-9230

I ERIM/ SAR Ivan Cindrich system reviewBDM Mike Haney (1st week of Apr.)

May 1994UI Dynetics Pulse Compress. Neil Mohon CECOM ESM demo

(at CECOM) 205-922-9230

I

g Dynetics Pulse Compress. Neil Mohon final review(Huntsville) 205-922-9230

December 1994

ERIM/ SAR Ivan Cindrich system review/demoBDM Mike Haney (1st week of Dec.)

mU

II

mmOctober 20, 1992 4

l