|
From: Al D. <al...@ie...> - 2003-09-11 04:26:01
|
On Wednesday 10 September 2003 06:10 pm, Willis Shih wrote: > There is a similar simulator, xsim, which does multi-level > simulation. Its format seems to be the same as your gnucap. > Do you know what the difference between xsim and your gnucap > is? I assume you mean the one published by Xing Zhou at Nanyang Technological University. It is a fork of a predecessor to Gnucap. He and I were grad students at U. of Rochester at the same time. His research was on monte-carlo methods for device modeling. He did a program "EMCUR", in FORTRAN. His contribution to XSIM is mostly in modeling and automatic correlation between analog and digital modes. The rest of it is mine. He has not shared any of it with me. He has had a good academic environment to work in. I have never had that. I might now, time will tell. The whole treatment of "implicit mixed mode", the BBD matrix, and some of the benchmarks are straight out of my dissertation, or earlier presentations I did in grad school. A year or so ago, Paolo Nenzi asked him for a copy, and he wouldn't send it. Strictly, he must release source under GPL. I have never pressed it. The current version of Gnucap adds a model compiler, use of queues instead of bypass for latency and dormancy exploitation, and bunch of other things. Performance is better than the early version he forked to make XSIM. On Wednesday 10 September 2003 06:10 pm, Willis Shih wrote: > BTW, how can I get your Ph.D. dissertation? Could you mail > me a soft copy? I put a copy at: http://www.freeelectron.net/al-davis-dissertation.pdf Looking back at it .... There is a lot left out. |