ray@ifh.ethz.ch (Ray Ballisti) (10/24/88)
Dear colleagues, you will be surprise to get this letter . Well, I got your address from the group of Prof. Anliker. I asked to give me contact with people using Transputer or Quadputer. We are just now beginnig to work with those units. Our experience is so far positive. We have two Quadputer from MicroWay,with 8 INMO S T800. A few problems arise: the link daughterboard which interconnects the four T800 on the board is very sensitive (bad contacts), so that we got to return one u nit. Then the FORTRAN compiler does not provide COMPLEX *16 (DOUBLE PRECISION COMPLEX), which is very unconvenient for us, as we got to rewrite all the complex operations in such a way that we do it in DOUBLE PRECISION as REAL variables. Also we miss Graphic Tools. Our group ( called Electromagnetics Group) in the Dept. of FIELDS AND WAVES is working on programs to solve electromagnetic fields problems. We are applying a numerical method called MMP (Multiple MultiPole) which was developed in our grou p from an idea of Ch. Hafner, my colleague, as early as 1980. We are now able to sol ve quite complicated problem not only on a mainframe but also on a PC!! The raison of my writing is to get in touch with people using Quadputer in order to exchange information about algorithm, compiler bugs, connection problems between T800's, usw. I would be very glad if you could tell me what going on in this field at Cornelle University, especially for numerical computation. Thank you very much in advance for your answer! Ciao Ray