next up previous contents index
Next: Multi-Threaded Loading Up: Introduction to WRspice Previous: Types of Analysis   Contents   Index


Multi-threading

WRspice can use multi-threading to accelerate certain types of analyses. This capability is not present in the original Berkeley SPICE and most derivatives, and allows WRspice to take advantage of the presence of multiple processor cores provided in modern microprocessor chips. Multi-threading allows different cores to work on the same simulation job in parallel.

Multi-threading in WRspice is a new feature currently being incorporated. It should be considered somewhat experimental at this point. Users are encouraged to try it and share their observations (and report bugs!). Multi-threading does not guarantee a faster run, as it has its own overhead that must be overcome to reduce overall run time. Experience will provide insight into which types of circuits and analyses benefit from multi-threading and which do not, and what related threading parmeter values, such of the number of threads to reserve, give the best results on a given machine.

All supported operating systems provide multi-threading, however parallel runs require multiple cores or CPUs. Many Intel processors provide two instruction queues (threads) per core, so that the number of available hardware threads is twice the number of cores.



Subsections
next up previous contents index
Next: Multi-Threaded Loading Up: Introduction to WRspice Previous: Types of Analysis   Contents   Index
Stephen R. Whiteley 2022-09-18