AS-2

 

ACHE NOTES Gio Wiederhold

x ia
A Summary of the ACME System November 11, 1966
‘(Pecsent tion ¢ given ab the ONR Con Ge

 

Psychobiology Confercnes Ley 17,
the U.S. Novy Postgeeduete Seon,
Monberey, Cs “Litornia)

 

The ACMS ccmputer system at the Stanford Medical School is designed to
provide powerful computing to research leboratorics in the Medical School.
The type of computation service planned is regular batch processing (mainly
at night) end real time interective capability for on line experiments.

The remainder of this description will concern itself with the
jJmolementelion ef this lebter facilis Uys

fn order to accamodaate 11 the laboratories in the Medical School with
their widely verying data rates, the system is designed to share the aves.
able computer tine. Tae eaouns of time allotted to e user, hoveve ry is not
one Sixed unit peer period, i.e. 200 ms every 1O seconds, bat xrether the
time requircd by him to ecquivre end process one date poling.

 

the date can be obtedied in e munber of ditferent WOYS (sec eyppendin
CN-2, SD)

Date nay ve typed on a typewriter station. ALL textual and progremning
infomietion is entered this way.

Slow to modiwa speed (vp to 1000 samples/sec cond) analog signals can be
entered vic a cubsiditny conputer which can either sean input voltage
at the 1000/second rate or respond to separate interrupt signals given by
“the expe aeaueHll or experimenter. “he precision of the conversion cen be
up to Wi bits (. 02%) preaise. This computer will convert, reformat end
prepeocess these date end then transmit then to the main computer,

 

 

Ler to the
ers tneat
actunlly generate

Digitel input vp to 16 bits wide can be hendled in a mnennex :
aneloy drmpat with the subs idiery computer, in order to serve us
have their own digitising cqupaent. or Whose

 

 

    

 
   
   
 
 

Users oF % pith ivweh higher rete demands; especially those that
ary
Cu

 

currently conpeters installed, will be able to connec dincetly
to the me high spced (wp to 125 ,000 seaples per second)
pa. WL8 case they hy progress the - CR

f yn

 

PE? IOV ET » tut beac the roselte vo

    

gs
nm ch as daqevioping oc device to - elloy mediuna to high speed
response eccuisition vie rultiflexor ehannels: |

 

Re a ision of AS.
AS-2

Page 2

The IBM 360 Model 50 which we are wing ~ as the main computer has been
chosen with a configuration that will support this type of multiple user
activity. Its main (cove) memory Size is one million bytes ox chavacters,
or 250,000 words or values. The backup storage is an RM ple file, which
Stores deta on strips of magnetic tape, 2000 of them, which are all in-
dividually retrievable within 0.6 Seconds. It has a total capacity of loo
million bytes. The subsidiary computer is an TBM 1800 process

control computer, connected via a speciél direct channel to the main con-
puter,

Results may be listed on the system printer, of course, but the emphasis
will be on typing oul the results on the typewriters in the laboratories,

Data may also be returned in analog or digital form via the 1800 at rates
comparable to those of the input capability; generally within « few seconds
after the results have been produced inside the Model 50. ‘The digital lines
may be used to drive plotters producing graphical surmeries of the experi~
ments. Display equizment of various types can be installed in the lebora-
torics and be driven from the enalog or digital 1800 outputs.

the small. computers can elso be used to distribute output via their ty pe~
writer, display tubes and plotters, or they may use the results to auto-
nabically control the continnation of the experiments.

It is quite obvious that a system of the described scope is not supported
by any computer manufacturer, We have therefore designed and written e
Simple but complete Support packege including en interactive compiler, &
supervisory systen, input-output procedures and data acquisition end dis-
tribution routines. The system design is such thet continuous guidance is
Hel). to the user via the typewriter (see ACME note RC-1 and appendix
RU-1).

To make such en approach at all economically possible, we are programming
using an IRM conpiler (FORTRAN H) to alloy us to write the system within
the allotted time spen. This should also make it possible to later share
the results of our software development work with others. The lenguace
thet our system will compile and which we hope will be acceptable to ow
non- computer specialist users is e subset of Piel, a& FORTRAN like proce-
dure] language, defined by IBM for use on 360 systems. Thus we expect that
procedures checked and proven on our interactive compiler can be filled and
used om ovr and other computer libreries under standard systems (see

ACME note on PL-l: PL-1).

The input-output system will include file handling and retrieval facili-~
ties where ell deta filed will be eutonatically lebellled with ell perti-
nent infornation to optimize the usefulness of the collected infoxmetion.

No eleborete buffering procedures will be used to wine
tive tiiac; “the tine thet he eannot use Lor CO

 

§,

toned eves to tho next veer in the queue.
AS-2

Page 3

Since for some experiments the system reaction tine will te quite critical,
we will have to limit the nunaber of thes iu & piven period, An
attempt to utilize every availeble eos weer cycle Tor this work can onlzy
resvlt Jn system overloading and fas SVEN, & Mucor of users wilthh
non-critical problens, rowbsine processing or information retrieval can
balance the system to achieve reasonable total woilisation, and we plan
also to provide facilities for this type of use es soon es the more critin
cal uses are y ,

     

   
 

 

 

oO

The project currertly uses several differerit computers around Stenford to

check oul parts of the system and is preparing to do & siimlation study of
Pp &

the queving elgorithu. A small technical scoup is building prototype inter.

connection equipment for the various data interfaces,

The planning work was Sponsored by 4 Josiah Macy Foundation plonning grant and
further funding has been granted fron NIN end Macy Ioundation. Much credit for idecs
and procedures pocs to other computer instellations end other peop. ,

notably project MAC eat M.I.@., MEDLAB at the Latter-Day Saints Hospital in

Salt Leake City, the University of California at Berkeley Caxputetion Centex

énd ARPA project, U.C. San Francisco medical school, U.C.i.A. Hee
Sciences, ete., end of course the Computetion Center and the Cony
Science Depertuent of Stenford itsel<,