aymanrasheed7 / NetOpt

A hardware accelerated solution for topological network optimization problem: given a set of resistors and a resistance value, find a series-parallel combination of the resistors such that the equivalent resistance is as close as possible to the given resistance value.

NetOpt

General command

outFile ARG1 ARG2 ARG3 OPTIONALARG4 OPTIONALARG5

Argument examples

ARG1: "E12" (using {1.0, 1.2, 1.5, 1.8, 2.2, 2.7, 3.3, 3.9, 4.7, 5.6, 6.8, 8.2}), "ONE", "INT", "ODD"
ARG2: "E", "PI", "PHI", "SQRT" (sqrt(nR)), "SQRT2.5", "2.5", "-2.5" (2.5 with optional exclution)
ARG3: "1", "2", ..., "12" (nR = number of resistors)
OPTIONALARG4: "1", "2", ..., "12" (table size controller)
OPTIONALARG5: "1", "2", ..., "256" (number of threads to create)

Example usage

Windows

g++ .\NetOptThread.cpp; .\a.exe E12 SQRT 10

Or

g++ -lopencl .\NetOptGPU.cpp; .\a.exe E12 SQRT 10

Linux

g++ -pthread NetOptThread.cpp; ./a.out E12 SQRT 10

Or

g++ -pthread -lopencl NetOptGPU.cpp; ./a.out E12 SQRT 10

Example output

Windows

Solution: (1.2+(3.9|(2.2+(((1.0|(1.5+4.7))|(2.7|5.6))+(1.8|3.3)))))
Result: 3.1622776652262714 (845658744/267420775)
Target: 3.1622776601683795
Cost: 5.057892e-09
CPU time (s): 12.819000000
Execution time (s): 12.815042700

Linux

Solution: (1.2+(3.9|(2.2+(((1.0|(1.5+4.7))|(2.7|5.6))+(1.8|3.3)))))
Result: 3.1622776652262714 (845658744/267420775)
Target: 3.1622776601683795
Cost: 5.057892e-09
CPU time (s): 135.711018000
Execution time (s): 8.269306404

About GPU

Performance on GPU was worse than that on CPU. To run on GPU proper drivers are needed.

Tested using OpenCL 3.0 CUDA driver on NVIDIA GeForce RTX 3050 Ti Laptop GPU in Windows.

Did not work when using OpenCL 3.0 NEO driver on Intel(R) Iris(R) Xe Graphics in Windows.