SECTION I DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE FIJBIC HULM SERVICE APPLICATION FOR RESEARCH GRANT LEAVE BLANK EXCEPT FOR GRANT NUMBER NPE PROGRAM GRANT NUMBER 5 PO7 FR 0031142 RMEW GROUP FORMERLY CONTlNUATlON SUPPORT I DATE RECEIVED INVENTlO,, STATEMENT 0 YES 0 NO TO BE COMPSETED BY PRINCIPAL INVESTIGATOR ABEREVVITED l7llE OF RESEARCH PROJECT (SOL* 01 rhom M 1.J tit& of lbvarch Grwt Awa&dJ Advanced Computer for Medical Research 2 DATES OF ENTRE APPROVED PROJECT PERIOD 3. DATES OF NEK, BUDGET PERlOD /Uwol/y II monfhtl 4. AMT. REQUESTED FOR NEXT BUDGE: PERIOD ,D,w Cain Only)--Last IInn, Pooe 2 FROM THROUGH FROM THROUGH g-l-66 g-30-69 8-1-67 7-31-68 ' 52O,q3 SA NAME OF PplKlPM INVESTIGATOR (Lort. FinI, In~hal) Lederberg, Joshua H. MAIUNG ADDRESS OF PRINCIPAL INVESTIGATOR (Str&, City, 3.k ZP caa) Department of Genetics D. AREA CODE AND TELEPHONE NO.(S] `?I 7 Stanford University School of Palo Alto, California $304 ADDRESS WHERE RESEARCH WILL BE CONDUCTED (If some OS hem 5H) chskbo.rU Medicine Chairman, Computer Policy Committee Stanford University School of Medicine F. DEPARTMENT, PRYICE, LABORATORY OR EQUIVALENT Stanford University School of Medicine Palo Alto, California 94304 G. MUOR SUBDIMK)N Stanford University School of Medicine 7. ME FEDERAL FACILITIES TO BE USED FOR THIS RESEARCH~ m NO 0 YES X OF TIME [If yn, l;,f kcilitier and indkok cd& of VI, on CO"h"u.ahO" poser., TO BE COMPLETED BY RESPONSIBLE ADMINISTRATIVE AUTHORITY 8. APPuwT ORGANIZATION /Nams and A&&heel, Cib, Skre. z!p C&J Stanford University Stanford, California 94305 9. NAME. TITLE At4D ADDRESS OF OFFICIAL TO WHOM CHECKS SHOULD BE MAlLED Mr. K. D. Creighton, Controller Encina Hall, Stanford University Stanford, California 10. NAME AND TITLE OF OFFICIAL SIGNING FOR APPLICANT ORGANl7ATiON 11. IDENTIFV ORGANlZATlONAL COMPONENT RESFONSIBLE FOR CONDUCl OF SCIENTIFIC ASPECTS OF PROJECT (Murr be compkkd. Set lowuctionr , School of Medicine 12. TYPE OF ORGANIZATION (Check appk.ble ,hmJ 0 INDIVIDUAL PlJELlC INSTITUTION: 0 FEDERAl O STATE UL- IJ OTHER PRtVATE INSIITUtlON: B NONPROFIT 0 PROFIT 13. PHS ACCOUNT NUMBER (Eota 11. ESTABLISHED INDIRECT COST RATE (see l"dr"dra.lsJ 55 xsiw 15. TERMS AND CONDITIONS. The undersigned accept, os to any grant awarded, the obligation to comply with Public Health Sekice Research Project Grant Regu- lations in effect at the time of the award (42 CFR, Part 52) and the terms and conditions in the Grants for Research Projects Policy Statement. The undersigned further agree to comply with Title VI of the Civil Rights Act of 1964 (P.L. 88.3521, ond the Regulation issued pursuant thereto and state thot the formally filed Assurance of Compliance with such Regulation (Form HEW-441) applies to this project. The undersigned also certify that they have no commitments or obligations, Including those with respect to inventions, inconsistent with compliance to the above. 16. A. YGNATURE OF PRINCIPAL INVESTIGATOR DATE SIGNARURRS 549/67 L (SAME A5 PERSON NAMED IN lTEM 10 DATE SECTION II FROM THROUGH GRANT NUMBER SECTION H-BUDGET (USUALLY 12 MONTHS) 8-l-67 7 -31-68 FR 00311-02 ,. \TEM,ZE DIRECT COSTS FOP NEXT BUDGET PER03 PERSONNEL TIME OR SALARY FRINGE BENEFITS EFFORT WHRS REQUESTED TOTAL NAME (tart. Fan,, ,,&a,) TITLE OF POSITION (see l"It"Chm, ldl bl Ibt (4 (4 If) PRINCIPAL INVESTIGATOR See attached schedule SUPPLIES Computer Office Engineering $20,000 4,000 20,000 Note 1 * 44,000 DOMESTIC TRAVEL Note 2 * 4,000 - I FOREIGN I I INPATIENT HOSPITALIZATION HOSPITAUZATIO(*I (Study pahentr) OUTPATIENT COSTS -I----- ALTERATIONS ANO RENOVATIONS PUBLlCATlON COSTS Note 3 s 8, ooo ALL OTHER EXPENSES (See htrYChOnrJ Communication (telephone, p ostage) .s 2,576 -- TOTAL DIRECT COSTSiEnter on Page I, hem 4) $ 520,073. I Continuation page Name Lederberg, Joshua Wiederhold, Gio Breitbard, Gary Crouse, Linda Cummins, David Flexer, Jules Hintz, Gertrude Holtz, Klaus Moore, Mabel Patel, Arunkant Sanders, William ----mm- ------e Class, Charles ------- ------a Osborne, DeWape Weatherby, Albert Zilka, Teresa 323 00311-02 Personnel Budget August 1, 1967 - July 31, 1968 Title of Position Principal Investigator Associate Director Systems Programmer Systems Programmer Systems Programmer Engineer (2 mos.) Systems Programmer Engineer Statistician Systems Programmer Systems Programmer User Education Statistician Operations Supervisor Computer Operators Machinists Computer Technician Computer Technician Part time assistants Secretary Time or Effort 106 100 100 100 100 100 100 100 100 100 40 100 100 37 l/2 37 l/2 37 l/2 Salary Requested $ 16,538 12,540 9,720 13,608 2,330 10,368 12,960 6,930 9,396 14,040 3,920 9,600 9,639 27, ho* 1,800 5,990 4,992 4,800 5,928 Administrative Assistance by Stanford Computation Center 5,456 Total Salaries $187,795 Fringe Benefits (10.5%) %i'18 t-operators computed at 24 hour day x $3.38 per hour (av. operator wage) x 365 days per year x 1.15 (for holidays sick leave, and vacation) x .80 (ACME Share) $207,513 Continuation page FR 00311-02 Equipment Budget Equipment Rental DM 360/50 System Less 2% Stanford Campus Facility support subtotal IBM 360/50 DM 360/67 Utilization IBM 2741 Communication Terminals (20) IBM 1360 Disk Packs (12) $12/mo. Data Sets (10) rental@ $3O/mo. installation ,g $60. Rented lines (10) @ $ll.O5/mo. installation @ $15. IBM 1800: additional costs IBM 1826 @ $5gg/mo. IBM 1442 @ 212/mo. IBM 029 @ 62/mo. IBM 7720 @ 125/mo. maintenance @ lTl/mo. State sales tax on rented equipment, except data set installation, rented lines and installation, and 1800 maintenance subtotal, rented equipment $252,984 Note 4 $232,439 <46,488> 185,951. 20,000 16,800 3,600 600 1,326 1.50 $222,751 1,728 4,200 1,476 14,028 8,801 HS-398 (REV. 6-64) page 2-2 SECTION II (ConGnued) SECTION II-BUDGET (Continued) Grant Number m 00311-02 6. Supplemen!al information regarding ITEMS in the proposed budget for the next period which requre explonohon. See Instruchons. (Use Continuation Pager if necessary.) (1) Due to the higher utilization of the equipment supplies for engineering and shopwork are higher than expected in the original application ($44,000 rather than $35,000). (2) The amount requested fro domestic travel is less than half of the amount requested for the year 1966/1967 because of the lessened need to collect information and ideas from other centers. Due to various delays travel was actually largely paid for by the planning grant provided by the Jos. Macy, Jr. Foundation. The requested amount will support professional travel to seminars and meetings of direct professional value, (3 > We expect during the coming year to complete the system and make it and the relevant documentation available to other installations facing similar problems. C. If budget proposed in this submission represent; o significant change (either increase or decrease) from the LEVEL originally recommended for this budget period, (1) provide on explanation for the change, and (2) indicate the effect that this change will hove, if ony, on the conduct of research and general level of support for the remainder of the approved project period. (Use Continuoticn Poges if necessary.) NO CHANGE D. Designate items contained in the detailed budget presented on poge 2 for which indirect costs WILL NOT be claimed, or ore subject to negotiation. [Use Contlnuo- tion Pages if necessary.] NONE Continuation page m 00311-02 SECTION II - BUDGET (4) The delays that are being experienced with the IBM 360/67 time-sharing system lessen the benefits of plication.(Section 7, equipment page 12) sharing that were hoped for in the original ap- But since the utility of the 360/67 to the medical school is also diminished the net effect of this delay is negligible. 360/67 time is needed, however, to continue system development concurrent with regular time sharing operations. PHS-396 (REV. 6-641 b3e 3-l SECTION III SECTION MI-FISCAL DATA FOR FROM CURRENT BUDGET PERIOD (USuALLY I2 MOfmHS) Y/1-/66 ziJ7YJSPHs FR oo311.-ol I I I The following ipertains to your CURRENT budget. This Informotlon in coniunctlon wtth that provided on Page 2 wtll be used in determining the amount of additIonal support for the NEXT budget period. *. BUDGET CATEGORIES Personnel (Salaries) Fringe Benefits ConsultcxS Services Equipmenft Supplies -k Hospitalimtion Outpatieai? Costs Alteraticwus and Renovations Publication Costs Other Tel., Postage Camp. Time (Babcock Total Dirc+t Costs Indirect Cc&s TOTALS -y Use below space to: CUaRENT BUDGET II) 77,675 49,612 7,301 5,209 260,997 10,577 207,721 218,298 -42,699 15,250 . 18,716 15,000 33,716 _a8,466> 4,000 900 1,000 1,900 2,100 29,300 29,300 ----- 29,300 ----_ 2,750 2,200 2,340 401,813 42,721 444,534 ACTUAL EXPENDITURES THRU 3-31-67 (Insert Dote, 12) 556 1,500 4,382 2,056 694 4,652 2,000 6,382 < 4,182> 3,000 7,652 < 5,312> 123,904 27,287 151,191 ESTIMATED ADDITIONAL EX- PENDITURES FOR REMAINDER OF CURRENT BUDGET C rnoJE:YpD TOTAL ESTIMATED EXPENDITURES (Cd 2 plvs Cd 31 PI ESTIMATED UNEXPENDED BALANCE (Subhct Cal 4 from Cd. I, (51 ~8,140 _ 4,005 87.552 9,214 (I 0; VT> < 1,913> 272,366 396,270 5,543 20,977 ' 293,343 48,264 444,534 - 6. Explain any significant bolonces shown in column (5). C. List all other research support pending or approved not previously reported. (Identify by source, project title, and amount.] NONE B. The reduction in equipment rental resulted from delay in delivery. This was offset by an even longer delay in delivery of the software which resulted in increased project personnel to provide the basic programming system. Supplies ran higher than estimated because it was advantageous to purchase the componerm'ts and have the technical staff fabricate equipment rather than purchase certain items as: switching equipment, pulse generator and power supplies, equipment rack. Terminal services were leased from Allen Babcock Company to serve both the develop- ment of the ACME system and to provide educational service to the medical school in the areas of terminal programming. PHS-2590.1 Rev. 7.65 (All pevlaur edctionr obrolete) PAGE 4 (Use Confinval~on Page- 0% necerrory) SECTION IV APPLICANT REPEAT GRANi NUM2ER SHOWN ON PAGE I -* I GRANT NUMBER SECTION IV SUMMARY PROGRESS REPORT PROGRAM DIRCTOR {tort. FM. Instic,,) Lederberg, Joshua NAME OF ORGANI*ATION m 00311-02 PEKOD COvERED BY THIS REPORT FROM THROUGH Stanford University School of Medicine TITLE OF PROJECT (Reped We rhonr, sn item 1 on firrt pogej 8-1-67 7-31-68 - Advanced Computer for Medical Research 1. Clrt any publtcotions pendlng or pubhshed and not prewously reported. 2. List and describe any foreign trowel undertoken during the above Deriod 3. Dercnbe accomplishments ;~nce lost summary progress report. Specify and describe the significance of any changes in the direction token by the proje:t during the above period. During the report period the research effort was devoted almost exclusively to the development, programming, and check-out of the ACME computer system. Tkhe staff of the project was occupied with a number of tasks which will be reported in separate paragraphs. These will cover: System design and implementation Program translator Data acquisition and distribution programming Improvements in terminal man/machine interaction Data transmission tests and design Production and installation of basic transmission units Design of a CRT display unit for medical data Development of a basic statistical and text processing library for interactive use Consultation with medical staff and faculty Education for medical staff and faculty Installation of the computer equipment Operation of the computer facility The progress in most areas has been extremely satisfactory. In the short time between receipt and check-out of the basic IBM equipment (December 15, 1966) and this date (April 15, 1967), we have been able to transfer our systems work from the variety of computers used for development and to begin offering a very limited but true timesharing service to the medical school. The education program has led to a great deal of enthusiasm on the part of the medical staff and faculty. The greatest lack as of this date is the delay in file capability. We can expect to have this capability available by June 1, 1967. During this project much assistance and exchange has been effected between ACME and the Central Campus Facility at S-f;anford. With the current delays in soft- ware delivery of IBM's timesharing system the ACME System will be used at the Central Facility. Continued exchange is going on between ACME and the Stanford Linear Accelerator Center in the areas of high speed data transmission and graphic support and between ACME and the Allen-Babcock Company regarding terminal control and systems develop- ment. In addition, there has been a joint educational and informational effort in the 1800 computer area with the Syntex Laboratories of Palo Alto. Throughout the ensuing paragraphs references are made to the working papers of the ACME Project, or ACME Notes, which are appended to this report as Appendix I. Continuation page 273 00311-02 SECTION IV - SUMMARY PROGRESS REPORT (continued) Interest in the ACME System has been wide-spread. Presentations have been invited and given at: I. II. III. IV. v. VI. VII. VIII. IX. UCLA Health Sciences Computing Facility ONR Workshop on Psycho-biology & Computers (Naval Post-Grad School, Monterey, California) The paper has been published in the proceedings of the conference. 3/18/66 ?/17/66 IBM Research in Cambridge, Mass. IEEE Workshop on Progress in Time Sharing Lancaster, Pa. 6/10/66 10/12/66 University of Toronto, Toronto, Canada Argonne National Laboratories Conference on Time-Sharing Model 50's The paper is due to be published as part of the proceedings. 8/4/66 ll/1/66 COMMON Meeting, New Orleans, La. u/29/66 A report has been published through COMMON prior to this meeting which is the result of a joint effort of IBM and four 1800 users, including ACME: Report of the 1800 Time Sharing Executive System Review Committee. ACM, Los Angeles Section 2/l/67 The presentation has been reviewed in the March issue of DATAMATION. CALTECH, Pasadena 3/29/Q In addition, a system description has appeared in IB14's internal locumentation. Even though our system is barely operational, IBM has made arrangements mder which the University of Witwatersand, Johannesburg, South Africa and the Jniversity of Paris, France have visited us, and are currently communicating extensively on the applicability of the ACME System for their installations. The progress experienced would not have been possible without the enthusiastic support of the staff. Continuation page FR 00311-02 NAME Gio Wiederhold .Gary Y. Breitbard Charles Class Linda Crouse (part-time) .David E. Cummins 'Robert Flexer .Ann Hintz Klaus Holtz Zeva LaHorgue (part-time) Gerald Miller (leave of absence) Mabel Moore Arun Pate1 William S. Sanders Voy Wiederhold (part-time `1 !?IOR ACME STAFF HIGHEST YEARS OF PREVIOUS POSITION DEGREE EXPERIENCE OR EXPERIENCE BS 1957 MA 1-963 a 1963 I29 units BA 1963, MS 1965 BA 196s BS lg& BA 1958 BA 1964 BS 1965 MA 1965 MS 1964 MA 1960. In addition to the secretary there are three 9 Visiting Prof., IIT, Kanpur, India Head of Programming, U.C. Berkeley Wrote Student Compiler at U.C. Berkeley Operator, Stanford Desert Research Institute U.C., Davis Programmer, IBM Post-graduate work, U.C., Berkeley Electrical Engineering Associate Systems Engineer Associate Programmer, IBM Engineer at Linear Accelerator, Stanford State Dept. of Public Health, Berkeley Bureau of Chronic Diseases Systems Programmer III, U.C.) Berkeley Genetics Dept., Stanford Electrical Engineering, U.C. Berkeley Time sharing project group General Electric, Phoenix Programmer, UCLA, Health Sciences technician/operators on the full-time staff and a number of part-time student assistants, *Charged to funds provided by the Josiah Macy, Jr. Foundation. HS-396 (REV. 6-641 page 5-2 Continuation page m 00311-02 PERSONNEL DISTRIBUTION SENIOR STAFF CLERICAL MONTHLY OR L & JUNIOR CONTRACT GW GYE CC LC DEC RF AH KH ZL iMM AP WJS VW WITH S.C.C. 4anagement,Adm. .5 1 1 jystem Dev. .5 1. .2 1 Compiler Dev. .2 1. .8 )ata acquisition .4 .8 .7 2.5 program lardware Dev. 1. .5 lardware Prod. .4 3.3 %atistics,etc. .l .4 .7 lonsultation .2 .3 .l .2 .l .l .3 .l :ducation .4 bperations 14 3 1.5396 (REV. 6-64) page 5-3 Continuation page m 00311-02 1. SYSTEM DESIGN AND DEVELOPMENT Thanks to the additional funding provided by the Josiah Macy, Jr. Foundation we had been able to do design, simulate, and test hypotheses for the ACME computer system prior to the grant period. I A simulation, written in Burroughs Algol and run on the Stanford Computation Center B5500 computer (ACME Note &L-l) indicated that the proposed scheduling algo- rithm using "yielding" - logic rather than "time-slice-cutoff" logic was valid, For the simulation data from the MIT timesharing system (MAC) (ACME Note OP-1) and the large-computer experience from the Stanford and U.C. Berkeley 7090/7094 systems were used to insure a system responsive to qualitatively large demands. This yielding-logic is a radical departure from large timesharing systems (ACME Note AY-1) currently in existence and promises to give the facilities and computing power required by serious researchers. One system using this logic currently in a less general environment is the MEDLAB system at the LDS Hospital, Salt Lake City, Utah, set up by Dr. Homer Warner. Another aspect of a system in a life-science area is that the data tends to be voluminous. To aid the researcher in the problem of keeping track of the amount of data that can be handled by a large computer system a formal method for handling data is being implemented. All data stored by the system can be automatically identified with the user's name , project, the date and time, file sequence numbers and the actual data name (ACME Notes FC-1, FI-3, FD-1, FF-1, FP-1, FE-l). Indexes to the data are kept separately to facilitate updating, safekeeping, and search for data previously collected (ACME Notes FA-1, FLU-l). Storage modes are limited to textual data and real type numeric data, IBM support for the mechanicalstorage device (the IBM 2321 pie file) did not arrive until February 15, 1967, and a con- siderable percentage of the ACME effort is now going into the provision of the file capability. The remainder of the system is operational and is able to support the hardware maximum of 15 users at typewriter terminals, quests. (ACME Notes IOA-2, 10-2, MA-l) and to respond to real time data re- The majority of the system is written in FORTRAN, including the scheduling and file supervisory mechanisms so that this system can be transformed to operate on other equipment with a minimumof effort. It also enables us to adjust the system easily on the basis of data gathered from experience, to accommodate equipment change:, and to introduce new ideas and procedures. I 2. THE PROGRAM TRANSLATOR The ACME/PL translator is a true incremental compiler. As far as we can deter- mine it is the only one currently in existence, even though much discussion on the usefulness of such a tool in the timesharing environment has taken place. It generates absolute binary machine language code from the PL/l statements typed in by the user. Through careful subsetting of the PL/l language (ACME Note PL-2) we can deliver to the user a quality of computational power which is otherwise limited to batch systems. The compiler is operational, although language extensions continue to be added. This rapid development was again made possible largely due to the fact that the predecessor to this compiler, the STUDENT compiler at the Berkeley Computation Center IBM 7040-7094 system had been written in FORTRAN; that planning funds were made available by the Josiah Macy foundation; that the Stanford 7090 system and the IBM 360-50 computers at the Stanford Linear Accelerator and at the Allen-Babcock Company were made available for testing and development (ACME Notes HL-2, HDC-1 and YA-1). PM-398 IREV. 6-64) page 5-4 I Continuation page m 00311-02 1 Initial measurements of the performance of the generated code indicate a slightly higher execution speed than IBM's PL/l compiler currently achieves, al- though that compiler runs in a batch mode only, and considerably higher speeds than interpretive systems as are generally in use now in interactive time-sharing systems (ACME Notes KO-6, MT-2, ND-l, NT-b, NS-1). The compiler takes care also of command handling, (ACME Notes RC-1, LA-2, PC-2) and user debugging facilities (ACME Notes RU-2, IX-3) and is designed to present to the user a consistent one-level interface, The user input/output facilities are not yet completed, but available are the facilities we expect to be used generally: free format multi-value input in response to system prompts and system formatted multi- value output (ACME Notes FS-2, OF-l, OP-1, TV-2). I 3. DATA ACQUISITION AND DISTRIBUTION PROGRAMMING The programming to control realtime data acquisition and distribution has not yet been integrated into the system. facility was operating smoothly. This work could not commence until the computer We initially experienced considerable difficulty in our efforts to keep information transfer compatible with IBM's Operating System Standards. This goal required intensive and extensive study of the system (ACME Notes CP-2, OS-2, CL-l, DN-1, DG-2). We have overcome this hurdle now and we are very satisfied with IBM design in this area. A change of IBM support staff recently has helped us also in the communication required with the manufacturer. We are not able to com- municate properly with non-IBM supported devices as PDP-8's, line-computers, dis- plays, and data transmission apparatus, The integration of these programs into the time-sharing system is scheduled to begin May 1, 1967, is expected in June. and limited user availability In the meantime, check-out of hardware and some production data acquisition is taking place on a scheduled non-time-sharing basis using the software in its current status. I 4. MAN-MACHINE INTERACTION In order to improve the man-machine interaction ACME is equipping its terminals with an indicator panel (ACME Note CI-2) so that the user is always aware of the status of the computer and his problem, In addition an ATTENTION Key has been pro- grammed (ACME Note PA-2) to give the user ability to immediately interrupt whatever action the system or his program is carrying out. Another aspect of satisfying the responsiveness of the system is the data acquisition implementation, which assures that the user controls when his equipment is to be sampled (ACME Note HA-l). I 5. DATA TRANSMISSION TESTING AND DESIGN Considerable effort has been made in testing and developing means for economic data transmisstion throughout a modern building such as the Stanford Medical School (ACME Note HDT-1). This has led us to connect our IBM terminals to the IBM computer directly. The use of telephone communication equipment is hereby avoided with an attendant cost reduction. The cost reduction is largely due to the fact that data transmission use of voice telephone facilities, mind, is less than ideal. which are engineered with another set of problemsin An ACME built switching panel is located at the computer operator's console for connecting users into the system (ACME Notes KA-1, KB-2, HSW-1 From remote locations we use data-phone and voice frequency FM coded data trans- mission. Continuation page m 00311-02 ACME has to thank the Instrumentation Research Laboratory of the Department of ienetics who are sharing both their facilities and experience to make the current Level of work possible. 5. HARDWARE PRODUCTION AND INSTALLATION As a result of the field work detailed above, standard circuits have been designed and built. The majority of these used integrated circuit logic, and are nounted and checked out within the ACME Facility (ACME Note m-1). These are now in stock as off-the-shelf items and can be combined with standard power supplies in a rack mounted units containing the required number and types of data input and output ievices. Two labs (Respiration and Pediatric Cardiology) are currently routinely transmitting data via these links to ACME and the ACME Engineering staff is working st the installation of additional links (ACME Note HT-1). An input/output typewrite: has been connected to one input line to give lower case alphabetic keypunching cap- ability (ACME Note KP-1). A link to a small computer has recently been checked out and others are being assembled. Much of this effort is being done on a cost sharing basis with the laboratories involved, both to conserve ACME funds, and to insure joint real interest in the pro- jects. I LIST OF TERMINALS REQUESTED TO DATE NAME Dr. WassermaqVan Kessel & Dr. Luetcher Dr. Bagshar! Dr. T. Merigan Dr. George Wertheim Ed Brown Dr. J. W. Bellville Dr. K. M. Colby Dr. K. L. Chow Dr. E. Dong Dr. Djerrassi Dr. W. Forrest Dr. Fred Fox Dr. Allen Gates Dr. A. Goldstein Dr. D. Harrison Dr. L. Herzenberg Dr. A. M. Iannone Dr. K. Killam Dr. Kopell Dr. J. Lederberg Dr. S. Liebes Dr. Mesel/Conn Dr. Mesel/Northway Dr. Mesel/Radiology Dr. T. Nelson W. Reynolds Dr. L. Rosenberg Dr. Stewart (3) Dr. L. Stryer (2) Dr. S. Kountz J. Hwang `M-398 IREV. 6-64) DEPARTMENT DATE OF INSTALLATION Dept. of Medicine Anesthesia Computer Science Dept. Neurology Surgery Genetics/Chemistry Anesthesia Clinical Lab. Gynecology Pharmacology Cardiology Genetics Neurology Pharmacology Psychiatry, VAH Genetics Genetics Pediatrics Pediatrics Pediatrics Surgery Genetics Med. Micro-Biology Fleischman Lab. Biochemistry Dept. of Medicine Dept. of Medicine Radiology Infectious Diseases Psychiatry 12/20/66 l/16/ 67 12/20/66 l/16/67 l/16/67 12/20/66 2/6/67 1/ '16/67 Surgery l/16/67 Genetics Page 5-6 Continuation mae FR 00311-02 I 7. CATHODE RAY TUBE DISPLAY In order to present data quickly in the high data rate interactive experiments, cathode ray tube displays are of great value. The data to be displayed in much life-science work has the form of time series graphs, annotated with the results of the computer analysis. With this in mind, as a joint project of ACME and the Instrumentation Research Laboratory a CRT display has been designed with the following features: 1) Display tube is driven by digital logic to insure stability of display 2) The digital logic is controlled by an independent memory to give an economic source for the required regeneration cycles. 3) A memory organization is oriented toward vector display to allow effective use of the unit for time-series graphs, The proto-type of this unit is currently under test, being driven by FORTRAN programs in the 360 and has demonstrated the feasibility of the approach, The cost of the parts has been about $6,500 of which the majority is accounted for by the CRT tube itself and the core memory unit. The connection to the computer follows ACME small computer conventions and the programming logic is similar to the driving of the CALCOMP digital plotters so that the same programs may be used, I 8. STATISTICAL AND TEXT PROCESSING LIBRARY A beginning has been made with the development of a library to process data on this interactive system. A number of statistical highly interactive routines are currently available on ACME's Babcock terminal (ACME Notes EB-1, EBA-1, EBB-l, EBD-1, EBL-l), while a survey has been made of existing statistical routines (ACME Note ES-l) which are candidates for inclusion in the system. Testing of various of these is currently in progress, whereas about a dozen are currently available to the users on the Babcock terminal. Much experience is being gathered to organize these routines so that they may be used directly by the medical researcher without having to consult professional or semi- professional programming staff. It is hoped that by the end of the summer a fairly complete statistical library will be available and that the efforts of the group can then be diverted more to the problems of analyzing continously arriving data. As a by-product and extension of the compiler development some text processing routines have become available. These are not yet integrated into the timesharing system, but are available on a stand-alone basis. The current capabilities include text sorting (ACME Notes KC-l, KH-5) concordance preparation, word list with fre- quency count generation, key word in context type indexing and capability for spec- ifying uninteresting words for deletion (stop words). In the process of check-out are options for searching-for sentences-containing-specific-words, text comparisons, and generation of data for further statistical processing.. The routines are oriented toward the processing of large files and economic usage of core memory. (ACME Note WTXT-1) They have been used at Stanford for analyzing Rorschach test responses, psychiatric diagnosis and setting up clinic appointments. 9. CONSULTATION A fair amount of staff time has been spent in discussing with staff and faculty of the medical school the feasibility and approach to a large number of projects. ACME has gathered a good impression of the range of problems that the system will have to respond Co, and also found a few that cannot be solved with current technol- ogy and facilities. As a result of those discussions we are certain that the medical school will quickly saturate every resource that can be made available through the ____- PM-398 (REV. 6-641 system at its present leP981 of support. 5-7 Continuation page FR OOJjll-02 I 10. EDUCATION TO educate the medical faculty and staff in a manner that is directly related to their problems is one of the tasks of a specialized medical facility. To enable an early start for this area a terminal to the Allen-Babcock time-sharing system is being rented through the Stanford Computation Center. A monthly seminar is being conducted to inform the medical school of the pro- .gress with the Project, and to give us the opportunity to hear speakers from other institutions discuss their work in relevant areas. During November through February a series of 15 four-and-a-half hour courses were conducted which were successfully completed by 167 members of the medical school faculty and staff. (ACME Notes ABC-l thru 10) The current demand for these terminals, exceeds their availability to the extent that weekly sign-up is required, despite this having only limited computing capability and no data acquisition facilities. A user's manual (ACME Note AM-l) has just been drafted which again is oriented towards solving medical research problems. A new series of courses is due to start in May using ACME's own facilities. Ill. EQUIPMENT INSTALLATION The installation of the computer equipment in the specially built structure was finally completed on April 8, 1967. Due to delays in approval of the various funds used to construct a special structure and adapt it to the computer's requirements , primitive and novel means were used to enclose the computer and keep it operating while the construction progressed, The dust has finally settled down and the facility is now not only functional, but also extremely attractive and much commented on by visitors to the STANFORD-PALO ALTO HOSPITAL and the Medical School. Thanks are due here to the Stanford Business School who made a computer floor available, the medical school architect's office who believed that the impossible was possible, and IBM who were willing to risk their equipment to the elements and the construction crews, 12.. COMPUTER OPERATIONS The operation of the computer is handled through arrangements with the Stanford Central Facility which enables us to secure reliable 24-hour-T-day staffing without having to employ redundant back-up staff. The technical aspect of operations is supervised by a member of the ACME staff. Full staffing has resulted in very reliabl operations under unfavorable conditions and our machine has shown an availability during the period from 12/12/66 to 4/09/67 of 4.5 per cent after scheduled main- tenance (4.9). Much of the computer's time is being used for ACME hardware check- out. The ACME System development takes priority currently, but users problems are routinely run at least overnight. The system has a very poor batch performance since the requirements for time-. sharing have taken precedence in both hardware and software selection. {ACME Notes CN-2, CQ-1, DL-1, OD-1) It will be interesting to compare cost to productivity ratios when the time-sharing service is in full swing, Gio Wiederhold PHS-398 (REV. 6.641 Page 5-8