New program
vfAnalyst lets you write an efficient and effective multicore program that is designed for parallelism. This is useful both for desktop/server application programs and for embedded software.
First you write a simple “skeleton” program that only contains key functions and data structures. Nothing needs to work correctly at this point; the only important thing is to have key data structure writes and reads in place, even if they read and write only nonsense data.
This lets vfAnalyst show you how these functions will interact and whether any dependencies will block parallelization. You can also annotate critical functions with timing information (actual and target) to help establish timing budgets.
You can then parallelize this structure, eliminating blocker dependencies as necessary, to provide your program template. Your report will help you structure the code. From there you can start filling in the functions or assigning them to a team to fill in.
As more code is written, you can reanalyze to confirm that your timing is still on track and that no new blocker dependencies have been introduced.
vfSoftware will add even more capabilities for automating the design of application programs.
Future tools will add even more capabilities for automating the design of embedded systems.
What do you want to do next?
- See a brief video showing what vfAnalyst can do.
- Take a quick tour of vfAnalyst.
- Start a free 30-day vfAnalyst trial.
- Read a whitepaper about how vfAnalyst depicts program behavior.
- Read a whitepaper about how to parallelize with vfAnalyst.
- Learn more about the Vector Fabrics cloud implementation.
- Learn about pricing.
- Learn more about Vector Fabrics.