Commit Graph

5393 Commits

Author SHA1 Message Date
Holger Vogt 7ec2708aaa Setting up compilation with CUDA 10.1 and Visual Studio 2017 2019-09-07 19:33:05 +02:00
Holger Vogt 3af4a867e0 Update to the how-to-... 2017-10-08 15:49:20 +02:00
Holger Vogt eed79f70c9 update for use with cude 2017-10-08 15:17:54 +02:00
Holger Vogt 19f8d18c23 remove OpenMP entries. OpenMP is not supported. 2017-10-08 15:17:31 +02:00
Holger Vogt cbe2da9d64 remove unused 2017-10-08 15:14:29 +02:00
Holger Vogt fe9b8b13be Remove XSPICE from spinit 2017-10-08 15:00:20 +02:00
Holger Vogt 97e4e027f2 Make search path for spinit relative to executable. 2017-10-08 14:59:28 +02:00
Holger Vogt e688436b4d remove old project files which are not supported anymore. 2017-10-08 14:58:56 +02:00
h_vogt ebdc9ea7ab no OpneMP support
DEbug info only when debug version
use /LTCG for optimized code
2017-09-24 00:57:13 +02:00
h_vogt f8782139c0 enhance compute capability beyond deprecated values:
compute_50,sm_50
debug info only when debug version is compiled
2017-09-24 00:57:13 +02:00
rlar 678fdd7f16 fixup 2017-09-23 21:48:05 +02:00
h_vogt 31c25db5a7 A tiny instruction of compiling ngspice with CUDA
using Visual Studio 2015
2017-09-23 21:40:17 +02:00
h_vogt 6a920e73e1 VS project files for CUDA 2017-09-23 21:24:15 +02:00
h_vogt b31d29ffe5 add a missing definition of CONSTvt0 2017-09-23 21:22:45 +02:00
h_vogt ccb3631a8d add mssing definition of M_PI 2017-09-23 21:21:31 +02:00
h_vogt 0db7a24d0d config.h, disable XSPICE and CIDER 2017-09-23 21:20:36 +02:00
Francesco Lannutti 0d4c76fd5d Added the support for CUDA and non-CUDA models within the same netlist in CUSPICE
Added KLU support for '.ic' and '.nodeset' instructions, needed also by CUSPICE
2017-09-20 19:36:33 +02:00
Francesco Lannutti 5c766ed162 Let NGSPICE overwrite Pivoting Tolerance within KLU with PivRel 2017-09-20 19:36:25 +02:00
Francesco Lannutti d19e2faea2 Add the support for '.ic' and '.nodeset' instructions within KLU 2017-09-20 19:36:22 +02:00
Francesco Lannutti 108e1f510c Relaxed ABSTOL and pushed GMIN to make c7552 circuit convergence 2017-09-17 17:24:14 +02:00
Francesco Lannutti d05cdc2c0b Fixed a bug in CKTload regarding CKTdiagGmin. It where added 2 times 2017-09-17 17:24:14 +02:00
Francesco Lannutti a8241785fe Fixed again the wrapper for libtool for Linux 2017-09-17 17:24:14 +02:00
Francesco Lannutti c47f0432f4 Fixed configuration for MacOS 2017-09-17 17:24:14 +02:00
Francesco Lannutti dc91caed79 Added new architectures and fixed a bug for MacOS (Darwin) 2017-09-17 17:24:14 +02:00
rlar e0e6176ba1 cuda test, a bsim dc i(vd) plot 2017-09-17 17:24:14 +02:00
rlar ce4e1493bd cuspice testcases, fix inductor values 2017-09-17 17:24:14 +02:00
rlar 7df02eb234 gtx1050ti with debian nvida-cuda-toolkit 7.5 2017-09-17 17:24:14 +02:00
rlar 6eef531172 bsim4 (4v8), hisimhv2, initialize cuDEVdestroy and DEVtopology slots 2017-09-17 17:24:14 +02:00
rlar 7030c1d2bf automake, version 15, introduce subdir-objects
for CUSPICE/*.c files in a subdirectory
2017-09-17 17:24:14 +02:00
Francesco Lannutti 0f659520f1 Fixed a bug in CKTtrunc reduction when there is nothing to reduce 2017-09-17 17:24:14 +02:00
Francesco Lannutti 6b134116b2 Moved Truncation Error Calculation into GPU for CUSPICE 2017-09-17 17:24:13 +02:00
Francesco Lannutti 0c89a0be3a Fixed CUSPICE compilation under NVCC versions prior than 6.0. CUSPICE will be compiled for every architectures supported by NVCC 2017-09-17 17:24:13 +02:00
Francesco Lannutti 72614c8f48 Fixed a bug in CUSPICE 2017-09-17 17:24:13 +02:00
Francesco Lannutti b52f33b629 Fixed a bug in libtool_wrapper_for_cuda.tcl to handle correct compilation of -fPIC objects 2017-09-17 17:24:13 +02:00
Francesco Lannutti 4dbc588b31 Whitespaces and cosmetic changes to improve readability 2017-09-17 17:24:13 +02:00
Francesco Lannutti 1afbccdb3b CUSPICE Integration from scratch 2017-09-17 17:24:13 +02:00
Francesco Lannutti fbec617c84 Fixed the KLU binding for BJT 2017-09-17 17:24:13 +02:00
Francesco Lannutti 203513a33f Commented out some debug printings 2017-09-17 17:24:13 +02:00
Francesco Lannutti fa61e0b9ac Fixed the two routines for SPARSE (thx to Marcel Hendrix) 2017-09-17 17:24:13 +02:00
rlar 13b0c64383 (int) casts, applied to a double which just has been limited to INT_MAX 2017-09-17 17:24:13 +02:00
rlar 6ff8182e82 casting, use (int) sizeof(), to swallow some warnings 2017-09-17 17:24:13 +02:00
rlar de51032ff0 use TMALLOC 2017-09-17 17:24:13 +02:00
Francesco Lannutti 693e1a1b4f Handling the general case for an empty CSC matrix during the conversion to CSR format 2017-09-17 17:24:13 +02:00
Francesco Lannutti e91a551370 Ordered the COO format also along the columns to have a correct CSR format 2017-09-17 17:24:13 +02:00
Francesco Lannutti caf51cdddc Fixed the AC part of the Sensitivity Analysis 2017-09-17 17:24:12 +02:00
Francesco Lannutti 8cf82e9275 Fixed the CSC to CSR conversion for KLU matrix-vector multiplication 2017-09-17 17:24:12 +02:00
Francesco Lannutti b30743b4bc Fixed Sensibility Analysis for KLU - First Trial 2017-09-17 17:24:12 +02:00
Francesco Lannutti 0ad00ea15d Avoid crashing during the Sensitivity Analysis in KLU, stating that it isn't supported 2017-09-17 17:24:12 +02:00
Francesco Lannutti ec8e4369a8 Fixed Sensitivity Analysis integration for KLU 2017-09-17 17:24:12 +02:00
Francesco Lannutti 7f7025af8a Fixed the sign inversion calculation for the determinant of KLU 2017-09-17 17:24:12 +02:00